This project develops an activity recognition model for a mobile fitness app using statistical analysis and machine learning. By processing smartphone sensor data, it extracts features to train models that accurately recognize user activities.

Supervised Classfication models - Logistic Regression & Decision Tree, AUC-ROC Curve

Improving a Machine Learning Model

Data Project

Performed EDA, Data Pre-processing, Imbalance data and Supervised Machine learning to predict customer transaction is fraud using features such as services that customer has signed up for, customer account information, and demographic information about the customer.

A python GUI application that uses a Convolutional Neural Network built in Tensorflow and Keras to classify chest x-rays into NORMAL or PNEUMONIC. The model has been trained on the dataset obtained from Kaggle and produces a good recall score of 94% on the test set.

This project explores the Framingham Heart disease dataset with the objective to predict its risk in 10 years. Various methods for handling missing values and outliers are explored as iterations. After analysing the dataset, important and necessary features are selected. Seven ML models are implemented, with evaluation on the basis of Test Recall.

Classification problems with imbalance datasets, a SMOTE approach

This is the Data Mining Project for predicting the student's grade before the final and Mid-2 examination. I use Python and Jupyter Notebook for this Project.

Machine-learning models to predict credit risk using free data from LendingClub. Imbalanced-learn and Scikit-learn libraries to build and evaluate models by using Resampling and Ensemble Learning

An attempt to study various ML models for predicting the quality of Red Wine using various performance measures.

A machine learning project designed to predict the likelihood of heart disease based on a set of health indicators.