Kaggle Competitions and Projects

Overview

This repository contains a collection of data analysis and machine learning projects, including competition submissions from Kaggle.

1. Daily Temperature of Cities

• File: Daily Temp of Cities/daily-temp-of-cities.ipynb
• Description: Analyzes and visualizes daily temperatures of major cities.
  • Cleans data for outliers and missing values.
  • Plots the highest and lowest temperature increases from 1995 to 2019.
  • Maps average temperature differences globally for comparison.

2. Titanic Competition

• File: Titanic Competition/Titanic ML Submission.ipynb
• Description: Predicts survival on the Titanic using machine learning models.
  • Handles missing data in 'Age', 'Fare', and 'Embarked'.
  • Applies one-hot encoding to categorical features.
  • Trains RandomForest, XGBoost, KNeighbors, and GradientBoosting classifiers, stacking predictions with XGBoost for final submission.

3. Predict Future Sales Competition

• File: Predict Future Sales Competition/Predict Future Sales Submission 1.ipynb

• Description: Predicts total sales for the next month using machine learning.

  • Prepares data by dropping unnecessary columns and training a RandomForestRegressor model.
  • Generates predictions and saves results to submission.csv.

• File: Predict Future Sales Competition/Predict Future Sales Submission 2.ipynb

• Description: Collaborative effort with Ajay Kumaar.

  • Cleans data by handling missing values, extracting date-related features, and encoding categorical variables.
  • Trains multiple classifiers (RandomForestClassifier, XGBClassifier, KNeighborsClassifier, SVC) and uses stacking for final predictions.

Data Sources

• Daily Temperature of Major Cities
• Titanic Data
• Predict Future Sales Data

Collaboration

• The Predict Future Sales Submission 2.ipynb notebook was a collaboration with Ajay Kumaar.

Usage

To explore these projects:

• Load respective data on Kaggle.
• Run the notebooks to reproduce the analyses and results.