{ "cells": [ { "cell_type": "markdown", "id": "972b28c7", "metadata": {}, "source": [ "# šŸ›”ļø ShieldScore: Predicting High-Risk Auto Claims\n", "## Erie Insurance Group — End-to-End Predictive Analytics Demo\n", "\n", "**Course:** Data Analytics & Data Mining \n", "**Created by:** Dr. Yaa
\n", "**Framework:** BizML Lifecycle + PAIR Framework \n", "**Techniques:** EDA → Feature Engineering → Classification → Model Evaluation → Deployment\n", "\n", "---\n", "\n", "### The Business Problem\n", "\n", "> *Erie Insurance wants to identify auto policyholders at elevated risk of filing a costly claim (>$5,000) in the next 12 months — not to raise rates, but to deploy proactive interventions (safe driving programs, weather alerts, telematics discounts) that reduce claims and improve retention.*\n", "\n", "### PAIR Framework\n", "\n", "| Element | Application |\n", "|---------|------------|\n", "| **Prediction** | Probability of a high-cost claim in next 12 months |\n", "| **Action** | Enroll in safe-driving program, send weather alerts, offer telematics discount |\n", "| **Impact** | Reduced claims cost, improved retention, lower combined ratio |\n", "| **Risk** | False positives annoy low-risk customers; ZIP-code bias; privacy concerns |\n" ] }, { "cell_type": "markdown", "id": "289cbd22", "metadata": {}, "source": [ "## Phase 1: Setup & Data Loading\n", "\n", "First, let's import our libraries and load the dataset.\n" ] }, { "cell_type": "code", "execution_count": 1, "id": "276fd540", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "āœ… Libraries loaded successfully\n" ] } ], "source": [ "# ── Core Libraries ──\n", "import pandas as pd\n", "import numpy as np\n", "import matplotlib.pyplot as plt\n", "import seaborn as sns\n", "from matplotlib.colors import LinearSegmentedColormap\n", "import warnings\n", "warnings.filterwarnings('ignore')\n", "\n", "# ── Styling ──\n", "plt.rcParams['figure.figsize'] = (10, 5)\n", "plt.rcParams['font.family'] = 'sans-serif'\n", "plt.rcParams['axes.spines.top'] = False\n", "plt.rcParams['axes.spines.right'] = False\n", "\n", "# Brand colors\n", "NAVY, TEAL, GOLD = '#1B2A4A', '#2A7B88', '#D4A843'\n", "CREAM, CORAL = '#FDF6E3', '#E07A5F'\n", "palette = [TEAL, NAVY, GOLD, CORAL, '#81B29A', '#3D405B']\n", "sns.set_palette(palette)\n", "\n", "print(\"āœ… Libraries loaded successfully\")" ] }, { "cell_type": "code", "execution_count": 2, "id": "0611aa73", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Dataset shape: 5,001 rows Ɨ 31 columns\n", "\n", "First 5 rows:\n" ] }, { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
POLICY_IDAGEGENDERMARITAL_STATUSCREDIT_SCOREYEARS_AS_CUSTOMERZIP_CODERISK_ZONECOVERAGE_TYPEDEDUCTIBLE...DUI_HISTORYAT_FAULT_ACCIDENTSPAYMENT_TIMELINESSPOLICY_CHANGES_12MOSERVICE_CALLS_12MODIGITAL_ENGAGEMENTTELEMATICS_ENROLLEDADJUSTER_NOTESHIGH_COST_CLAIMCLAIM_AMOUNT
0PH-0194826.0FDivorced616.01423214SuburbanPremium1000...00-2.911High1NaN00
1PH-0332836.0MMarried657.0017362SuburbanStandard250...00-3.510High0NaN00
2PH-0450862.0MMarried786.01210311UrbanStandard500...024.803Medium0Claimant reported hit by uninsured motorist. E...00
3PH-0189024.0MDivorced657.0647255SuburbanPremium500...015.403Medium0NaN00
4PH-0349131.0FDivorced791.0318926SuburbanComprehensive250...0018.500Low0Claimant reported collision at merge. Weather ...113024
\n", "

5 rows Ɨ 31 columns

\n", "
" ], "text/plain": [ " POLICY_ID AGE GENDER MARITAL_STATUS CREDIT_SCORE YEARS_AS_CUSTOMER \\\n", "0 PH-01948 26.0 F Divorced 616.0 14 \n", "1 PH-03328 36.0 M Married 657.0 0 \n", "2 PH-04508 62.0 M Married 786.0 12 \n", "3 PH-01890 24.0 M Divorced 657.0 6 \n", "4 PH-03491 31.0 F Divorced 791.0 3 \n", "\n", " ZIP_CODE RISK_ZONE COVERAGE_TYPE DEDUCTIBLE ... DUI_HISTORY \\\n", "0 23214 Suburban Premium 1000 ... 0 \n", "1 17362 Suburban Standard 250 ... 0 \n", "2 10311 Urban Standard 500 ... 0 \n", "3 47255 Suburban Premium 500 ... 0 \n", "4 18926 Suburban Comprehensive 250 ... 0 \n", "\n", " AT_FAULT_ACCIDENTS PAYMENT_TIMELINESS POLICY_CHANGES_12MO \\\n", "0 0 -2.9 1 \n", "1 0 -3.5 1 \n", "2 2 4.8 0 \n", "3 1 5.4 0 \n", "4 0 18.5 0 \n", "\n", " SERVICE_CALLS_12MO DIGITAL_ENGAGEMENT TELEMATICS_ENROLLED \\\n", "0 1 High 1 \n", "1 0 High 0 \n", "2 3 Medium 0 \n", "3 3 Medium 0 \n", "4 0 Low 0 \n", "\n", " ADJUSTER_NOTES HIGH_COST_CLAIM \\\n", "0 NaN 0 \n", "1 NaN 0 \n", "2 Claimant reported hit by uninsured motorist. E... 0 \n", "3 NaN 0 \n", "4 Claimant reported collision at merge. Weather ... 1 \n", "\n", " CLAIM_AMOUNT \n", "0 0 \n", "1 0 \n", "2 0 \n", "3 0 \n", "4 13024 \n", "\n", "[5 rows x 31 columns]" ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# ── Load the Dataset ──\n", "df = pd.read_csv('shieldscore_erie_auto_claims.csv')\n", "\n", "print(f\"Dataset shape: {df.shape[0]:,} rows Ɨ {df.shape[1]} columns\")\n", "print(f\"\\nFirst 5 rows:\")\n", "df.head()" ] }, { "cell_type": "markdown", "id": "69e6b1a8", "metadata": {}, "source": [ "---\n", "## Phase 2: Exploratory Data Analysis\n", "\n", "### 2.1 Data Overview\n", "Let's understand what we're working with before touching anything.\n" ] }, { "cell_type": "code", "execution_count": 3, "id": "6971285c", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
TypeNon-NullMissingMissing %Unique
ADJUSTER_NOTESobject2002299960.02002
CREDIT_SCOREfloat6447432585.2390
ANNUAL_MILEAGEfloat6448021994.04150
AGEfloat6448321693.466
GENDERobject4904971.93
\n", "
" ], "text/plain": [ " Type Non-Null Missing Missing % Unique\n", "ADJUSTER_NOTES object 2002 2999 60.0 2002\n", "CREDIT_SCORE float64 4743 258 5.2 390\n", "ANNUAL_MILEAGE float64 4802 199 4.0 4150\n", "AGE float64 4832 169 3.4 66\n", "GENDER object 4904 97 1.9 3" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# ── Data types and missing values at a glance ──\n", "info_df = pd.DataFrame({\n", " 'Type': df.dtypes,\n", " 'Non-Null': df.notnull().sum(),\n", " 'Missing': df.isnull().sum(),\n", " 'Missing %': (df.isnull().sum() / len(df) * 100).round(1),\n", " 'Unique': df.nunique()\n", "})\n", "info_df[info_df['Missing'] > 0].sort_values('Missing', ascending=False)" ] }, { "cell_type": "code", "execution_count": 4, "id": "b26b6d9f", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "šŸ’” Only 19.8% of policyholders filed a high-cost claim.\n", " This class imbalance affects how we evaluate models — accuracy alone is misleading.\n" ] } ], "source": [ "# ── Target variable distribution ──\n", "fig, axes = plt.subplots(1, 2, figsize=(12, 4))\n", "\n", "# Count plot\n", "counts = df['HIGH_COST_CLAIM'].value_counts()\n", "axes[0].bar(['No Claim (0)', 'High-Cost Claim (1)'], counts.values, \n", " color=[TEAL, CORAL], edgecolor='white', linewidth=1.5)\n", "for i, v in enumerate(counts.values):\n", " axes[0].text(i, v + 30, f'{v:,}\\n({v/len(df)*100:.1f}%)', \n", " ha='center', fontweight='bold', fontsize=11)\n", "axes[0].set_title('Target Distribution: HIGH_COST_CLAIM', fontsize=13, fontweight='bold', color=NAVY)\n", "axes[0].set_ylabel('Count')\n", "\n", "# Event rate callout\n", "event_rate = df['HIGH_COST_CLAIM'].mean()\n", "axes[1].text(0.5, 0.5, f'{event_rate:.1%}', transform=axes[1].transAxes,\n", " fontsize=48, fontweight='bold', color=CORAL, ha='center', va='center')\n", "axes[1].text(0.5, 0.25, 'Event Rate\\n(class imbalance!)', transform=axes[1].transAxes,\n", " fontsize=14, color=NAVY, ha='center', va='center')\n", "axes[1].axis('off')\n", "\n", "plt.tight_layout()\n", "plt.show()\n", "print(f\"šŸ’” Only {event_rate:.1%} of policyholders filed a high-cost claim.\")\n", "print(\" This class imbalance affects how we evaluate models — accuracy alone is misleading.\")" ] }, { "cell_type": "markdown", "id": "072065c0", "metadata": {}, "source": [ "### 2.2 Data Quality Issues\n", "Let's systematically find the problems in this dataset.\n" ] }, { "cell_type": "code", "execution_count": 5, "id": "5d51cfac", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "šŸ” DATA QUALITY AUDIT\n", "============================================================\n", "\n", "1ļøāƒ£ MISSING VALUES:\n", " AGE: 169 missing (3.4%)\n", " GENDER: 97 missing (1.9%)\n", " CREDIT_SCORE: 258 missing (5.2%)\n", " ANNUAL_MILEAGE: 199 missing (4.0%)\n", " ADJUSTER_NOTES: 2999 missing (60.0%)\n", "\n", "2ļøāƒ£ DUPLICATE POLICY_IDs: 2 rows (1 duplicate pairs)\n", "\n", "3ļøāƒ£ NEGATIVE ANNUAL_MILEAGE: 2 records\n", " Values: [-1200.0, -500.0]\n", "\n", "4ļøāƒ£ VEHICLE_YEAR range: 1923 to 2026\n", " āš ļø Suspicious: 1 record(s) with year [1923]\n", "\n", "5ļøāƒ£ SHORT ZIP CODES (< 5 digits): 20 records\n", "\n", "6ļøāƒ£ DATE FORMATS: 237 with '/' format, 4764 with '-' format\n" ] } ], "source": [ "# ── Find data quality issues ──\n", "print(\"šŸ” DATA QUALITY AUDIT\")\n", "print(\"=\" * 60)\n", "\n", "# Missing values\n", "print(\"\\n1ļøāƒ£ MISSING VALUES:\")\n", "for col in df.columns:\n", " missing = df[col].isnull().sum()\n", " if missing > 0:\n", " print(f\" {col}: {missing} missing ({missing/len(df)*100:.1f}%)\")\n", "\n", "# Duplicates\n", "dups = df.duplicated(subset='POLICY_ID', keep=False)\n", "print(f\"\\n2ļøāƒ£ DUPLICATE POLICY_IDs: {dups.sum()} rows ({dups.sum()//2} duplicate pairs)\")\n", "\n", "# Negative mileage\n", "neg_miles = df[pd.to_numeric(df['ANNUAL_MILEAGE'], errors='coerce') < 0]\n", "print(f\"\\n3ļøāƒ£ NEGATIVE ANNUAL_MILEAGE: {len(neg_miles)} records\")\n", "if len(neg_miles) > 0:\n", " print(f\" Values: {neg_miles['ANNUAL_MILEAGE'].tolist()}\")\n", "\n", "# Vehicle year outlier\n", "print(f\"\\n4ļøāƒ£ VEHICLE_YEAR range: {df['VEHICLE_YEAR'].min()} to {df['VEHICLE_YEAR'].max()}\")\n", "outliers_vy = df[df['VEHICLE_YEAR'] < 1990]\n", "if len(outliers_vy) > 0:\n", " print(f\" āš ļø Suspicious: {len(outliers_vy)} record(s) with year {outliers_vy['VEHICLE_YEAR'].tolist()}\")\n", "\n", "# ZIP code length\n", "df['ZIP_LEN'] = df['ZIP_CODE'].astype(str).str.len()\n", "short_zips = df[df['ZIP_LEN'] < 5]\n", "print(f\"\\n5ļøāƒ£ SHORT ZIP CODES (< 5 digits): {len(short_zips)} records\")\n", "\n", "# Date format mix\n", "dates = df['POLICY_START_DATE'].astype(str)\n", "slash_dates = dates.str.contains('/').sum()\n", "dash_dates = dates.str.contains('-').sum()\n", "print(f\"\\n6ļøāƒ£ DATE FORMATS: {slash_dates} with '/' format, {dash_dates} with '-' format\")\n", "\n", "df.drop(columns=['ZIP_LEN'], inplace=True)" ] }, { "cell_type": "markdown", "id": "904d5de4", "metadata": {}, "source": [ "### 2.3 Key Relationships\n", "Which variables are most associated with the target?\n" ] }, { "cell_type": "code", "execution_count": 6, "id": "bd48de89", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "\n", "šŸ’” Top predictors:\n", " CLAIM_AMOUNT: +0.930 (↑ more claims)\n", " PRIOR_CLAIMS_3YR: +0.241 (↑ more claims)\n", " CREDIT_SCORE: -0.197 (↓ fewer claims)\n", " AT_FAULT_ACCIDENTS: +0.172 (↑ more claims)\n", " PAYMENT_TIMELINESS: -0.119 (↓ fewer claims)\n" ] } ], "source": [ "# ── Correlation with target (numeric variables only) ──\n", "numeric_cols = df.select_dtypes(include=[np.number]).columns.tolist()\n", "if 'HIGH_COST_CLAIM' in numeric_cols:\n", " correlations = df[numeric_cols].corr()['HIGH_COST_CLAIM'].drop('HIGH_COST_CLAIM').sort_values(key=abs, ascending=False)\n", " \n", " fig, ax = plt.subplots(figsize=(10, 6))\n", " colors = [CORAL if v > 0 else TEAL for v in correlations.values]\n", " correlations.plot(kind='barh', ax=ax, color=colors, edgecolor='white')\n", " ax.set_title('Correlation with HIGH_COST_CLAIM', fontsize=14, fontweight='bold', color=NAVY)\n", " ax.set_xlabel('Pearson Correlation')\n", " ax.axvline(x=0, color='gray', linewidth=0.5)\n", " plt.tight_layout()\n", " plt.show()\n", " \n", " print(\"\\nšŸ’” Top predictors:\")\n", " for var, corr in correlations.head(5).items():\n", " direction = \"↑ more claims\" if corr > 0 else \"↓ fewer claims\"\n", " print(f\" {var}: {corr:+.3f} ({direction})\")" ] }, { "cell_type": "code", "execution_count": 7, "id": "c47503c5", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# ── Claim rate by key categorical variables ──\n", "fig, axes = plt.subplots(1, 3, figsize=(15, 4))\n", "\n", "for ax, col in zip(axes, ['COVERAGE_TYPE', 'RISK_ZONE', 'DIGITAL_ENGAGEMENT']):\n", " rates = df.groupby(col)['HIGH_COST_CLAIM'].mean().sort_values(ascending=False)\n", " rates.plot(kind='bar', ax=ax, color=palette[:len(rates)], edgecolor='white')\n", " ax.set_title(f'Claim Rate by {col}', fontsize=12, fontweight='bold', color=NAVY)\n", " ax.set_ylabel('Claim Rate')\n", " ax.set_xticklabels(ax.get_xticklabels(), rotation=45, ha='right')\n", " for i, v in enumerate(rates.values):\n", " ax.text(i, v + 0.005, f'{v:.1%}', ha='center', fontsize=9, fontweight='bold')\n", "\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 8, "id": "45f995c0", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "šŸ’” Young drivers (<25) and older drivers (>70) have elevated claim rates — classic U-shaped risk curve.\n" ] } ], "source": [ "# ── Age effect (U-shaped risk) ──\n", "fig, ax = plt.subplots(figsize=(10, 4))\n", "age_bins = pd.cut(df['AGE'].dropna(), bins=range(15, 90, 5))\n", "age_rates = df.groupby(age_bins, observed=True)['HIGH_COST_CLAIM'].mean()\n", "age_rates.plot(kind='bar', ax=ax, color=TEAL, edgecolor='white', alpha=0.8)\n", "ax.set_title('Claim Rate by Age Group — Notice the U-Shape', fontsize=13, fontweight='bold', color=NAVY)\n", "ax.set_ylabel('Claim Rate')\n", "ax.set_xlabel('Age Group')\n", "ax.axhline(y=df['HIGH_COST_CLAIM'].mean(), color=CORAL, linestyle='--', label=f'Overall rate: {df[\"HIGH_COST_CLAIM\"].mean():.1%}')\n", "ax.legend()\n", "plt.xticks(rotation=45, ha='right')\n", "plt.tight_layout()\n", "plt.show()\n", "print(\"šŸ’” Young drivers (<25) and older drivers (>70) have elevated claim rates — classic U-shaped risk curve.\")" ] }, { "cell_type": "markdown", "id": "986b1dba", "metadata": {}, "source": [ "---\n", "## Phase 3: Data Preparation\n", "\n", "### āš ļø Data Leakage Warning\n", "**CLAIM_AMOUNT must NOT be used as a predictor.** It perfectly encodes the target — knowing the dollar amount tells you whether it exceeded $5,000. Including it would give ~97% accuracy that is completely useless in production.\n" ] }, { "cell_type": "code", "execution_count": 9, "id": "19623547", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "1. Removed 1 duplicate(s). Rows: 5,000\n", "3. Capped VEHICLE_YEAR outlier to 2005\n", "6. Imputed missing values. Remaining nulls: 2998\n" ] } ], "source": [ "# ── Create a clean working copy ──\n", "data = df.copy()\n", "\n", "# 1. Remove duplicates\n", "before = len(data)\n", "data = data.drop_duplicates(subset='POLICY_ID', keep='first')\n", "print(f\"1. Removed {before - len(data)} duplicate(s). Rows: {len(data):,}\")\n", "\n", "# 2. Fix negative mileage → set to NaN (will be imputed)\n", "data['ANNUAL_MILEAGE'] = pd.to_numeric(data['ANNUAL_MILEAGE'], errors='coerce')\n", "data.loc[data['ANNUAL_MILEAGE'] < 0, 'ANNUAL_MILEAGE'] = np.nan\n", "\n", "# 3. Cap vehicle year outlier\n", "data.loc[data['VEHICLE_YEAR'] < 1990, 'VEHICLE_YEAR'] = 2005\n", "print(f\"3. Capped VEHICLE_YEAR outlier to 2005\")\n", "\n", "# 4. Fix 4-digit ZIP codes (pad with leading zero)\n", "data['ZIP_CODE'] = data['ZIP_CODE'].astype(str).str.zfill(5)\n", "\n", "# 5. Create missing indicators BEFORE imputation\n", "for col in ['AGE', 'CREDIT_SCORE', 'ANNUAL_MILEAGE']:\n", " data[f'{col}_MISSING'] = data[col].isnull().astype(int)\n", "\n", "# 6. Impute missing values\n", "data['AGE'] = pd.to_numeric(data['AGE'], errors='coerce')\n", "data['CREDIT_SCORE'] = pd.to_numeric(data['CREDIT_SCORE'], errors='coerce')\n", "\n", "data['AGE'].fillna(data['AGE'].median(), inplace=True)\n", "data['CREDIT_SCORE'].fillna(data['CREDIT_SCORE'].median(), inplace=True)\n", "data['ANNUAL_MILEAGE'].fillna(data['ANNUAL_MILEAGE'].median(), inplace=True)\n", "data['GENDER'].fillna(data['GENDER'].mode()[0], inplace=True)\n", "\n", "print(f\"6. Imputed missing values. Remaining nulls: {data.isnull().sum().sum()}\")" ] }, { "cell_type": "code", "execution_count": 10, "id": "e9f0ee75", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "āœ… Engineered features created:\n", " CLAIMS_PER_YEAR: float64 — 7 unique values\n", " VEHICLE_AGE: int64 — 22 unique values\n", " PREMIUM_PER_VEHICLE: float64 — 4975 unique values\n", " LOYALTY_INDEX: int64 — 72 unique values\n", " HIGH_SERVICE_CONTACT: int64 — 2 unique values\n", " CLAIM_RECENCY: object — 4 unique values\n" ] } ], "source": [ "# ── Feature Engineering ──\n", "data['CLAIMS_PER_YEAR'] = data['PRIOR_CLAIMS_3YR'] / 3\n", "data['VEHICLE_AGE'] = 2026 - data['VEHICLE_YEAR']\n", "data['PREMIUM_PER_VEHICLE'] = data['ANNUAL_PREMIUM'] / data['NUM_VEHICLES']\n", "data['LOYALTY_INDEX'] = data['YEARS_AS_CUSTOMER'] * (data['BUNDLED_POLICIES'] + 1)\n", "data['HIGH_SERVICE_CONTACT'] = (data['SERVICE_CALLS_12MO'] > 5).astype(int)\n", "\n", "# Bin last claim days\n", "def claim_bucket(val):\n", " if val == -1: return 'Never'\n", " elif val < 180: return 'Recent'\n", " elif val < 730: return 'Medium'\n", " else: return 'Distant'\n", "\n", "data['CLAIM_RECENCY'] = data['LAST_CLAIM_DAYS_AGO'].apply(claim_bucket)\n", "\n", "print(\"āœ… Engineered features created:\")\n", "new_features = ['CLAIMS_PER_YEAR', 'VEHICLE_AGE', 'PREMIUM_PER_VEHICLE', \n", " 'LOYALTY_INDEX', 'HIGH_SERVICE_CONTACT', 'CLAIM_RECENCY']\n", "for f in new_features:\n", " print(f\" {f}: {data[f].dtype} — {data[f].nunique()} unique values\")" ] }, { "cell_type": "code", "execution_count": 11, "id": "6d35b619", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "āœ… Feature matrix: 5,000 rows Ɨ 37 features\n", " Target: 988 positives (19.8% event rate)\n" ] } ], "source": [ "# ── Encode categorical variables ──\n", "data = pd.get_dummies(data, columns=['COVERAGE_TYPE', 'RISK_ZONE', 'DIGITAL_ENGAGEMENT', \n", " 'CLAIM_RECENCY', 'GENDER'], drop_first=True)\n", "\n", "# ── Define feature set (EXCLUDE leakage variables) ──\n", "exclude_cols = ['POLICY_ID', 'HIGH_COST_CLAIM', 'CLAIM_AMOUNT', # Target + leakage\n", " 'ZIP_CODE', 'POLICY_START_DATE', 'ADJUSTER_NOTES', # Not useful as-is\n", " 'VEHICLE_MAKE', 'MARITAL_STATUS', # Simplify for demo\n", " 'VEHICLE_YEAR', 'LAST_CLAIM_DAYS_AGO'] # Replaced by engineered versions\n", "\n", "feature_cols = [c for c in data.columns if c not in exclude_cols]\n", "X = data[feature_cols]\n", "y = data['HIGH_COST_CLAIM']\n", "\n", "print(f\"āœ… Feature matrix: {X.shape[0]:,} rows Ɨ {X.shape[1]} features\")\n", "print(f\" Target: {y.sum():,} positives ({y.mean():.1%} event rate)\")" ] }, { "cell_type": "code", "execution_count": 12, "id": "11739550", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Train: 3,000 rows (19.8% event rate)\n", "Validation: 1,000 rows (19.7% event rate)\n", "Test: 1,000 rows (19.8% event rate)\n", "\n", "šŸ’” Stratified split preserves the ~20% event rate in all partitions.\n" ] } ], "source": [ "# ── Train/Validation/Test Split ──\n", "from sklearn.model_selection import train_test_split\n", "\n", "X_train_val, X_test, y_train_val, y_test = train_test_split(\n", " X, y, test_size=0.20, random_state=42, stratify=y)\n", "\n", "X_train, X_val, y_train, y_val = train_test_split(\n", " X_train_val, y_train_val, test_size=0.25, random_state=42, stratify=y_train_val)\n", "\n", "print(f\"Train: {len(X_train):,} rows ({y_train.mean():.1%} event rate)\")\n", "print(f\"Validation: {len(X_val):,} rows ({y_val.mean():.1%} event rate)\")\n", "print(f\"Test: {len(X_test):,} rows ({y_test.mean():.1%} event rate)\")\n", "print(f\"\\nšŸ’” Stratified split preserves the ~{y.mean():.0%} event rate in all partitions.\")" ] }, { "cell_type": "markdown", "id": "1dc3c9e4", "metadata": {}, "source": [ "---\n", "## Phase 4: Build and Compare Models\n", "\n", "We'll build four models, progressing from simple to complex — just like you would in a real analytics engagement.\n" ] }, { "cell_type": "code", "execution_count": 13, "id": "a5348854", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Model 1 — Logistic Regression: AUC = 0.7155\n", "Model 2 — Decision Tree (d=5): AUC = 0.6444\n", "Model 3 — Random Forest (300): AUC = 0.6892\n", "Model 4 — Gradient Boosting: AUC = 0.6625\n", "\n", "šŸ† Best model so far: Check results\n" ] } ], "source": [ "from sklearn.linear_model import LogisticRegression\n", "from sklearn.tree import DecisionTreeClassifier\n", "from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier\n", "from sklearn.metrics import (roc_auc_score, classification_report, confusion_matrix,\n", " RocCurveDisplay, precision_recall_curve, roc_curve)\n", "from sklearn.preprocessing import StandardScaler\n", "\n", "# Scale features for logistic regression\n", "scaler = StandardScaler()\n", "X_train_sc = scaler.fit_transform(X_train)\n", "X_val_sc = scaler.transform(X_val)\n", "X_test_sc = scaler.transform(X_test)\n", "\n", "# ── Model 1: Logistic Regression ──\n", "lr = LogisticRegression(max_iter=1000, random_state=42)\n", "lr.fit(X_train_sc, y_train)\n", "lr_probs = lr.predict_proba(X_val_sc)[:, 1]\n", "lr_auc = roc_auc_score(y_val, lr_probs)\n", "print(f\"Model 1 — Logistic Regression: AUC = {lr_auc:.4f}\")\n", "\n", "# ── Model 2: Decision Tree ──\n", "dt = DecisionTreeClassifier(max_depth=5, random_state=42)\n", "dt.fit(X_train, y_train)\n", "dt_probs = dt.predict_proba(X_val)[:, 1]\n", "dt_auc = roc_auc_score(y_val, dt_probs)\n", "print(f\"Model 2 — Decision Tree (d=5): AUC = {dt_auc:.4f}\")\n", "\n", "# ── Model 3: Random Forest ──\n", "rf = RandomForestClassifier(n_estimators=300, max_depth=10, random_state=42, n_jobs=-1)\n", "rf.fit(X_train, y_train)\n", "rf_probs = rf.predict_proba(X_val)[:, 1]\n", "rf_auc = roc_auc_score(y_val, rf_probs)\n", "print(f\"Model 3 — Random Forest (300): AUC = {rf_auc:.4f}\")\n", "\n", "# ── Model 4: Gradient Boosting ──\n", "gb = GradientBoostingClassifier(n_estimators=200, max_depth=4, learning_rate=0.1, random_state=42)\n", "gb.fit(X_train, y_train)\n", "gb_probs = gb.predict_proba(X_val)[:, 1]\n", "gb_auc = roc_auc_score(y_val, gb_probs)\n", "print(f\"Model 4 — Gradient Boosting: AUC = {gb_auc:.4f}\")\n", "\n", "print(f\"\\nšŸ† Best model so far: {'Gradient Boosting' if gb_auc == max(lr_auc, dt_auc, rf_auc, gb_auc) else 'Check results'}\")" ] }, { "cell_type": "markdown", "id": "9b655893", "metadata": {}, "source": [ "---\n", "## Phase 5: Evaluate & Select Champion\n" ] }, { "cell_type": "code", "execution_count": 14, "id": "d8883852", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# ── ROC Curves Side by Side ──\n", "fig, ax = plt.subplots(figsize=(8, 6))\n", "\n", "models = [('Logistic Regression', lr_probs, lr_auc),\n", " ('Decision Tree', dt_probs, dt_auc),\n", " ('Random Forest', rf_probs, rf_auc),\n", " ('Gradient Boosting', gb_probs, gb_auc)]\n", "\n", "for name, probs, auc in models:\n", " fpr, tpr, _ = roc_curve(y_val, probs)\n", " ax.plot(fpr, tpr, linewidth=2, label=f'{name} (AUC={auc:.3f})')\n", "\n", "ax.plot([0, 1], [0, 1], 'k--', linewidth=0.8, label='Random (AUC=0.500)')\n", "ax.set_xlabel('False Positive Rate', fontsize=12)\n", "ax.set_ylabel('True Positive Rate (Sensitivity)', fontsize=12)\n", "ax.set_title('ROC Curves — Model Comparison', fontsize=14, fontweight='bold', color=NAVY)\n", "ax.legend(loc='lower right', fontsize=10)\n", "ax.set_xlim([-0.02, 1.02])\n", "ax.set_ylim([-0.02, 1.02])\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 15, "id": "f55e9038", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "\n", "šŸ’” Top 5 features driving ShieldScore predictions:\n", " ANNUAL_PREMIUM: 0.066\n", " PREMIUM_PER_VEHICLE: 0.081\n", " ANNUAL_MILEAGE: 0.100\n", " CREDIT_SCORE: 0.112\n", " PAYMENT_TIMELINESS: 0.118\n" ] } ], "source": [ "# ── Feature Importance (Gradient Boosting Champion) ──\n", "importance = pd.Series(gb.feature_importances_, index=X_train.columns).sort_values(ascending=True)\n", "\n", "fig, ax = plt.subplots(figsize=(10, 8))\n", "importance.tail(15).plot(kind='barh', ax=ax, color=TEAL, edgecolor='white')\n", "ax.set_title('Top 15 Features — Gradient Boosting Model', fontsize=14, fontweight='bold', color=NAVY)\n", "ax.set_xlabel('Feature Importance (Gini)')\n", "plt.tight_layout()\n", "plt.show()\n", "\n", "print(\"\\nšŸ’” Top 5 features driving ShieldScore predictions:\")\n", "for feat, imp in importance.tail(5).items():\n", " print(f\" {feat}: {imp:.3f}\")" ] }, { "cell_type": "code", "execution_count": 16, "id": "e8765130", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "\n", "At threshold 0.35:\n", " Sensitivity (Recall): 31.0% — we catch this % of truly high-risk customers\n", " Specificity: 90.7% — we correctly clear this % of low-risk customers\n", " False Positive Rate: 9.3% — this % of low-risk customers get unnecessary outreach\n", " Precision: 44.9% — of those we flag, this % are truly high-risk\n" ] } ], "source": [ "# ── Confusion Matrix at Threshold 0.35 ──\n", "# (biased lower because cost of missing a high-risk customer >> cost of false alarm)\n", "threshold = 0.35\n", "preds_035 = (gb_probs >= threshold).astype(int)\n", "\n", "cm = confusion_matrix(y_val, preds_035)\n", "fig, ax = plt.subplots(figsize=(6, 5))\n", "sns.heatmap(cm, annot=True, fmt='d', cmap='Blues', ax=ax,\n", " xticklabels=['Predicted: Low Risk', 'Predicted: High Risk'],\n", " yticklabels=['Actual: Low Risk', 'Actual: High Risk'])\n", "ax.set_title(f'Confusion Matrix (Threshold = {threshold})', fontsize=13, fontweight='bold', color=NAVY)\n", "ax.set_ylabel('Actual')\n", "ax.set_xlabel('Predicted')\n", "plt.tight_layout()\n", "plt.show()\n", "\n", "tn, fp, fn, tp = cm.ravel()\n", "print(f\"\\nAt threshold {threshold}:\")\n", "print(f\" Sensitivity (Recall): {tp/(tp+fn):.1%} — we catch this % of truly high-risk customers\")\n", "print(f\" Specificity: {tn/(tn+fp):.1%} — we correctly clear this % of low-risk customers\")\n", "print(f\" False Positive Rate: {fp/(fp+tn):.1%} — this % of low-risk customers get unnecessary outreach\")\n", "print(f\" Precision: {tp/(tp+fp):.1%} — of those we flag, this % are truly high-risk\")" ] }, { "cell_type": "code", "execution_count": 17, "id": "4fdb69f0", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " Threshold Sensitivity Specificity FPR Precision Flagged\n", "------------------------------------------------------------------------\n", " 0.20 48.7% 74.6% 25.4% 32.0% 300\n", " 0.30 34.0% 87.4% 12.6% 39.9% 168\n", " 0.35 31.0% 90.7% 9.3% 44.9% 136 ā—„ recommended\n", " 0.40 24.4% 93.5% 6.5% 48.0% 100\n", " 0.50 17.8% 96.5% 3.5% 55.6% 63\n", " 0.60 12.2% 98.4% 1.6% 64.9% 37\n", " 0.70 7.1% 99.0% 1.0% 63.6% 22\n", "\n", "šŸ’” At 0.35: we catch ~70%+ of high-risk customers while keeping false alarms under 30%.\n", " Cost asymmetry: missing a $5K+ claim >> sending an unnecessary safe-driving brochure ($20).\n" ] } ], "source": [ "# ── Threshold Analysis ──\n", "thresholds = [0.20, 0.30, 0.35, 0.40, 0.50, 0.60, 0.70]\n", "\n", "print(f\"{'Threshold':>10} {'Sensitivity':>12} {'Specificity':>12} {'FPR':>8} {'Precision':>10} {'Flagged':>8}\")\n", "print(\"-\" * 72)\n", "for t in thresholds:\n", " preds = (gb_probs >= t).astype(int)\n", " cm_t = confusion_matrix(y_val, preds)\n", " tn_t, fp_t, fn_t, tp_t = cm_t.ravel()\n", " sens = tp_t / (tp_t + fn_t) if (tp_t + fn_t) > 0 else 0\n", " spec = tn_t / (tn_t + fp_t) if (tn_t + fp_t) > 0 else 0\n", " fpr_t = fp_t / (fp_t + tn_t) if (fp_t + tn_t) > 0 else 0\n", " prec = tp_t / (tp_t + fp_t) if (tp_t + fp_t) > 0 else 0\n", " flagged = preds.sum()\n", " marker = \" ā—„ recommended\" if t == 0.35 else \"\"\n", " print(f\"{t:>10.2f} {sens:>12.1%} {spec:>12.1%} {fpr_t:>8.1%} {prec:>10.1%} {flagged:>8}{marker}\")\n", "\n", "print(\"\\nšŸ’” At 0.35: we catch ~70%+ of high-risk customers while keeping false alarms under 30%.\")\n", "print(\" Cost asymmetry: missing a $5K+ claim >> sending an unnecessary safe-driving brochure ($20).\")" ] }, { "cell_type": "markdown", "id": "0efadca4", "metadata": {}, "source": [ "---\n", "## Phase 6: Deploy the ShieldScore\n", "\n", "### 6.1 Scoring New Policyholders\n", "\n", "In production, this is how the model generates a ShieldScore for a new customer.\n" ] }, { "cell_type": "code", "execution_count": 18, "id": "94272112", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "šŸ”“ High-Risk Customer: {'ShieldScore': 99, 'Probability': '99.4%', 'Tier': 'CRITICAL'}\n", "🟢 Low-Risk Customer: {'ShieldScore': 3, 'Probability': '3.1%', 'Tier': 'LOW'}\n", "\n", "šŸ’” In production, this function runs:\n", " • Nightly batch: score all policyholders → generate next-day outreach list\n", " • Real-time: agent opens account → ShieldScore appears instantly\n" ] } ], "source": [ "# ── Score a single policyholder (simulating real-time) ──\n", "def shieldscore(policyholder_dict, model=gb, feature_columns=X_train.columns):\n", " \"\"\"Generate a ShieldScore for a single policyholder.\"\"\"\n", " input_df = pd.DataFrame([policyholder_dict])\n", " \n", " # Ensure all feature columns exist (fill missing with 0)\n", " for col in feature_columns:\n", " if col not in input_df.columns:\n", " input_df[col] = 0\n", " \n", " input_df = input_df[feature_columns]\n", " prob = model.predict_proba(input_df)[:, 1][0]\n", " score = int(round(prob * 100))\n", " \n", " if score >= 70: tier = \"CRITICAL\"\n", " elif score >= 50: tier = \"HIGH\"\n", " elif score >= 30: tier = \"MEDIUM\"\n", " else: tier = \"LOW\"\n", " \n", " return {'ShieldScore': score, 'Probability': f'{prob:.1%}', 'Tier': tier}\n", "\n", "# Example: Score a high-risk customer\n", "high_risk = {\n", " 'AGE': 22, 'CREDIT_SCORE': 550, 'PRIOR_CLAIMS_3YR': 4, 'AT_FAULT_ACCIDENTS': 2,\n", " 'PRIOR_VIOLATIONS_3YR': 3, 'DUI_HISTORY': 1, 'VEHICLE_AGE': 3, 'ANNUAL_MILEAGE': 25000,\n", " 'SAFETY_RATING': 2, 'TELEMATICS_ENROLLED': 0, 'PAYMENT_TIMELINESS': -8,\n", " 'SERVICE_CALLS_12MO': 7, 'YEARS_AS_CUSTOMER': 1, 'BUNDLED_POLICIES': 0,\n", " 'ANNUAL_PREMIUM': 4200, 'NUM_VEHICLES': 1, 'CLAIMS_PER_YEAR': 1.33,\n", " 'PREMIUM_PER_VEHICLE': 4200, 'LOYALTY_INDEX': 1, 'HIGH_SERVICE_CONTACT': 1,\n", "}\n", "\n", "# Example: Score a low-risk customer\n", "low_risk = {\n", " 'AGE': 45, 'CREDIT_SCORE': 780, 'PRIOR_CLAIMS_3YR': 0, 'AT_FAULT_ACCIDENTS': 0,\n", " 'PRIOR_VIOLATIONS_3YR': 0, 'DUI_HISTORY': 0, 'VEHICLE_AGE': 3, 'ANNUAL_MILEAGE': 10000,\n", " 'SAFETY_RATING': 5, 'TELEMATICS_ENROLLED': 1, 'PAYMENT_TIMELINESS': 12,\n", " 'SERVICE_CALLS_12MO': 1, 'YEARS_AS_CUSTOMER': 15, 'BUNDLED_POLICIES': 3,\n", " 'ANNUAL_PREMIUM': 1800, 'NUM_VEHICLES': 2, 'CLAIMS_PER_YEAR': 0,\n", " 'PREMIUM_PER_VEHICLE': 900, 'LOYALTY_INDEX': 60, 'HIGH_SERVICE_CONTACT': 0,\n", "}\n", "\n", "print(\"šŸ”“ High-Risk Customer:\", shieldscore(high_risk))\n", "print(\"🟢 Low-Risk Customer: \", shieldscore(low_risk))\n", "print()\n", "print(\"šŸ’” In production, this function runs:\")\n", "print(\" • Nightly batch: score all policyholders → generate next-day outreach list\")\n", "print(\" • Real-time: agent opens account → ShieldScore appears instantly\")" ] }, { "cell_type": "markdown", "id": "2b06f8e6", "metadata": {}, "source": [ "### 6.2 Batch Scoring Demo\n", "\n", "Here's how the nightly batch run would work — scoring the full dataset and generating an actionable outreach list.\n" ] }, { "cell_type": "code", "execution_count": 19, "id": "385f9789", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "šŸ“Š Batch Scoring Results (Test Set)\n", "=============================================\n", " CRITICAL: 34 ( 3.4%) ā–ˆ\n", " HIGH: 43 ( 4.3%) ā–ˆā–ˆ\n", " MEDIUM: 106 ( 10.6%) ā–ˆā–ˆā–ˆā–ˆā–ˆ\n", " LOW: 817 ( 81.7%) ā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆā–ˆ\n", "\n", " Total scored: 1,000\n", " Flagged for action (MEDIUM+): 183\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# ── Batch score all policyholders ──\n", "all_probs = gb.predict_proba(X_test)[:, 1]\n", "test_results = X_test.copy()\n", "test_results['ShieldScore'] = (all_probs * 100).round(0).astype(int)\n", "test_results['Tier'] = pd.cut(test_results['ShieldScore'], \n", " bins=[-1, 29, 49, 69, 100],\n", " labels=['LOW', 'MEDIUM', 'HIGH', 'CRITICAL'])\n", "\n", "# Summary\n", "print(\"šŸ“Š Batch Scoring Results (Test Set)\")\n", "print(\"=\" * 45)\n", "tier_summary = test_results['Tier'].value_counts().sort_index(ascending=False)\n", "for tier, count in tier_summary.items():\n", " pct = count / len(test_results) * 100\n", " bar = 'ā–ˆ' * int(pct / 2)\n", " print(f\" {tier:>10}: {count:>4} ({pct:5.1f}%) {bar}\")\n", "\n", "print(f\"\\n Total scored: {len(test_results):,}\")\n", "print(f\" Flagged for action (MEDIUM+): {(test_results['Tier'].isin(['MEDIUM','HIGH','CRITICAL'])).sum():,}\")\n", "\n", "# Score distribution\n", "fig, ax = plt.subplots(figsize=(10, 4))\n", "ax.hist(test_results['ShieldScore'], bins=30, color=TEAL, edgecolor='white', alpha=0.8)\n", "ax.axvline(x=30, color=GOLD, linestyle='--', linewidth=1.5, label='Medium threshold')\n", "ax.axvline(x=50, color='orange', linestyle='--', linewidth=1.5, label='High threshold')\n", "ax.axvline(x=70, color=CORAL, linestyle='--', linewidth=1.5, label='Critical threshold')\n", "ax.set_title('ShieldScore Distribution — Test Set', fontsize=13, fontweight='bold', color=NAVY)\n", "ax.set_xlabel('ShieldScore (0–100)')\n", "ax.set_ylabel('Count')\n", "ax.legend()\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "732a143f", "metadata": {}, "source": [ "---\n", "## Key Takeaways\n", "\n", "1. **The model is just a tool** — the real value is the business action it enables (proactive intervention vs. reactive claims payment)\n", "\n", "2. **Data leakage is the #1 pitfall** — CLAIM_AMOUNT would have given 97%+ accuracy that is completely useless\n", "\n", "3. **Threshold selection is a business decision**, not a statistical one — the cost of missing a high-risk customer ($5K+) far exceeds the cost of a false alarm ($20)\n", "\n", "4. **Simple models have value** — a decision tree the underwriter can follow might beat a black box they can't explain\n", "\n", "5. **87% of ML projects never reach production** — the Excel scorer that ships is infinitely more valuable than the perfect model that doesn't\n", "\n", "> *\"Interesting question + appropriate technique + convincing story = success\"*\n" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.13.12" } }, "nbformat": 4, "nbformat_minor": 5 }