Machine Learning and Data Visualizations
01
Loan Default Classification using Soft Voting Classifier
Built a Soft-Voting Classifier to predict loan defaulting customers; achieving an accuracy of up to 87%. Project includes data exploration, feature engineering and class sampling. We build a robust framework for tackling ML classification tasks. ⓘ
02
ANN Sentiment Analysis and Classification - Customer Reviews
A robust Deep Learning model to analyze restaurant reviews and output the overall sentiment in terms of number of stars. I undergo a data exploration and feature engineering process to develop suitable parameters for the ANN model. ⓘ
03
Number Plate Detection using OpenCV
I use Python and cv2 techniques such as Edge detection, Contour detection and Masking to extract number plate code from images and videos. ⓘ
04
Spotify Tracks API Data Engineering and Analysis
I use Python to build a pipeline to link the Spotify API, to a PostgreSQL database. Each line item is comprised of individual tracks and the features include track popularity, artist popularity, album release dates, etc. ⓘ
05
Stock Price Forecasting using RNN Deep Learning Model
Built a Recurrent Neural Network to forecast the TCS Stock Price. Using a step size of 60, the model is designed to forecast the opening price for the next 30 days. The model predicts the general flow of the market, eliminating the noise of frequent spikes and dips on an actual stock price chart and thereby giving the price action on a macro scale. ⓘ
05
NLP Project to Create Deep Learning Algorithm for Binary Classification
ANN NLP project to Analyze and Predict binary classification dataset. Project includes Data Exploratory and preprocessing designed to slim down number of parameters while minimizing impact on model performance. ⓘ
05
Inventory Tracking and Analysis Tool
Built an Inventory tracking and Analysis tool to better assist restaurant managers in tracking consumption rates, inventory value, cost of goods sold and demand over time to minimize loss due to theft and spoilage of goods. ⓘ
05
Kaggle Titanic Challenge
Project involves extensive feature engineering and analysis to maximize model performance and climb higher on the project leaderboard. Current accuracy stands at 78.2% ⓘ
05
COVID-19 Tableau Data Analysis Project
How has the Covid-19 Pandemic evolved in the year 2022? This is a Covid-19 Data Analysis project where I build a data pipeline using Python and use Tableau to storytell how the pandemic played out in 2022. ⓘ
