Add March Machine Learning Mania 2026 Kaggle submission notebook

March_ML_Mania_2026_Kaggle_Submission_Marktechpost.ipynb

@@ -0,0 +1,176 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# March Machine Learning Mania 2026 — Kaggle Submission Notebook\n",
+    "\n",
+    "**Competition:** [March Machine Learning Mania 2026](https://www.kaggle.com/competitions/march-machine-learning-mania-2026)\n",
+    "\n",
+    "This notebook generates a valid baseline submission for both the Men's and Women's 2026 NCAA Basketball Tournament predictions.\n",
+    "\n",
+    "> **Assumption:** The competition data files are available at `/kaggle/input/march-machine-learning-mania-2026/`.  \n",
+    "> We use `SampleSubmissionStage2.csv` as the template for the current-season submission."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Kaggle Notebook Code"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import os\n",
+    "import pandas as pd\n",
+    "\n",
+    "# ------------------------------------------------------------------\n",
+    "# Path to competition data (standard Kaggle input directory)\n",
+    "# ------------------------------------------------------------------\n",
+    "DATA_DIR = \"/kaggle/input/march-machine-learning-mania-2026\"\n",
+    "OUTPUT_DIR = \"/kaggle/working\"  # Kaggle saves output files here\n",
+    "\n",
+    "# ------------------------------------------------------------------\n",
+    "# Load the Stage 2 sample submission\n",
+    "# Stage 2 lists every possible 2026 team matchup that needs a prediction\n",
+    "# ------------------------------------------------------------------\n",
+    "sample_sub_path = os.path.join(DATA_DIR, \"SampleSubmissionStage2.csv\")\n",
+    "submission = pd.read_csv(sample_sub_path)\n",
+    "\n",
+    "print(f\"Sample submission shape: {submission.shape}\")\n",
+    "print(submission.head())"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# ------------------------------------------------------------------\n",
+    "# Validate the required columns are present\n",
+    "# The competition requires: ID (string) and Pred (float 0–1)\n",
+    "# ------------------------------------------------------------------\n",
+    "assert \"ID\" in submission.columns, \"Missing required column: ID\"\n",
+    "assert \"Pred\" in submission.columns, \"Missing required column: Pred\"\n",
+    "\n",
+    "print(f\"Required columns present: {list(submission.columns)}\")\n",
+    "print(f\"Total predictions required: {len(submission)}\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# ------------------------------------------------------------------\n",
+    "# Baseline model: predict 0.5 for every matchup (coin flip)\n",
+    "# This gives a valid first submission scored at 0.0 before tournaments\n",
+    "# Pred = probability that the LOWER TeamID team wins the matchup\n",
+    "# ------------------------------------------------------------------\n",
+    "submission[\"Pred\"] = 0.5\n",
+    "\n",
+    "print(\"Prediction column set to 0.5 (baseline) for all matchups.\")\n",
+    "print(submission.head(10))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# ------------------------------------------------------------------\n",
+    "# Quick sanity checks before saving\n",
+    "# ------------------------------------------------------------------\n",
+    "# Check all IDs follow the format SSSS_XXXX_YYYY\n",
+    "assert submission[\"ID\"].str.match(r\"\\d{4}_\\d{4}_\\d{4}\").all(), \\\n",
+    "    \"Some IDs do not match expected format SSSS_XXXX_YYYY\"\n",
+    "\n",
+    "# Check predictions are within [0, 1]\n",
+    "assert submission[\"Pred\"].between(0, 1).all(), \\\n",
+    "    \"Some predictions are outside the valid range [0, 1]\"\n",
+    "\n",
+    "# Confirm no duplicate IDs\n",
+    "assert submission[\"ID\"].is_unique, \"Duplicate IDs found in submission\"\n",
+    "\n",
+    "print(\"All sanity checks passed!\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# ------------------------------------------------------------------\n",
+    "# Save the submission file to the Kaggle working directory\n",
+    "# ------------------------------------------------------------------\n",
+    "output_path = os.path.join(OUTPUT_DIR, \"submission.csv\")\n",
+    "submission.to_csv(output_path, index=False)\n",
+    "\n",
+    "print(f\"Submission saved to: {output_path}\")\n",
+    "print(f\"File size: {os.path.getsize(output_path) / 1024:.1f} KB\")\n",
+    "print(\"\\nFinal submission preview:\")\n",
+    "print(submission.head())"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Kaggle Submit Command\n",
+    "\n",
+    "After the notebook finishes, submit the output file using the Kaggle CLI:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Run this cell to submit directly from the notebook (or paste into a terminal)\n",
+    "# Make sure your Kaggle API credentials are configured (~/.kaggle/kaggle.json)\n",
+    "\n",
+    "!kaggle competitions submit \\\n",
+    "    -c march-machine-learning-mania-2026 \\\n",
+    "    -f /kaggle/working/submission.csv \\\n",
+    "    -m \"Baseline 0.5 submission\""
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Notes\n",
+    "\n",
+    "- **File used:** `SampleSubmissionStage2.csv` — contains all 2026 season matchup IDs (both Men's and Women's).\n",
+    "- **Baseline:** Every matchup is predicted with probability `0.5` (equal chance for either team). This scores **0.0** before the tournament begins, as stated in the competition rules.\n",
+    "- **Pred column meaning:** Probability that the team with the **lower** `TeamID` wins. Men's TeamIDs are 1000–1999; Women's TeamIDs are 3000–3999.\n",
+    "- **Submission deadline:** March 19, 2026 4PM UTC. Select your best submission manually before the deadline — do not rely on automatic selection.\n",
+    "- To improve on the baseline, replace the `submission[\"Pred\"] = 0.5` line with model-predicted probabilities derived from the historical game data provided."
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "name": "python",
+   "version": "3.10.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}