This README provides an overview of various projects and homework assignments completed as part of my coursework in Applied Machine Learning and Data Mining. Each project is accompanied by a full report detailing the methodologies, tools, and outcomes.
- Problem 1: Regression analysis.
- Problem 2: Classification using K-Nearest Neighbors (KNN).
- Problem 3: Email spam detection.
- Problem 4: Binary image classification.
Second homework of my data mining course focusing on a binary classification problem: Includes all preprocessing steps, training, and validation. Optimization of K for KNN and Bayesian classification techniques.
An in-depth exploration of preprocessing steps for tabular, image, and text data:
- Tabular Data: Utilization of scikit-learn's pipeline class to compare and select preprocessing approaches.
- Image Data: First-time preprocessing.
- Text Data: Development of a Hamshahri corpus reader from scratch, text preprocessing with the Hazm library, and sentiment analysis using Polyglot.
First homework of my data mining course: Extensive methods for handling missing values, detecting outliers, and selecting features. Evaluation of methods using the accuracy of a neural network algorithm.
Work in Progress
Kaggle Competition Focus on predicting customer churn in the banking industry using PCA.
Kaggle Competition Project aimed at predicting obesity risk based on various factors with an achieved accuracy of 89%. Plans to refine and organize notebooks further.