Leveraging Big Data Analytics for Disease Classification Using Chatbot-Generated Features

Harness the power of Google Cloud Platform (GCP) to train diverse machine learning models-Random Forest, SVM and BERT and use their Pickled files for building sophisticated language model chatbot

Traffic Sign Classification

Implemented a high-impact Traffic Sign Classification model, leveraging Convolutional Neural Networks (CNNs) and Python. Employed KPIs to rigorously assess the model's effectiveness.

Gender Detection

Implemented a gender detection algorithm utilizing the kernel method of SVM, showcasing expertise in ml and statistical modeling. The algorithm was crafted, incorporating a polynomial kernel to enhance its discriminatory capabilities and then rigorously tested against carefully data.

Lost and Found System

Accuracy was heightened through LSTM recurrent neural networks for sentence encoding, and performance was optimized with text embedding and the Attention-GAN model. Additionally, a recommendation engine was seamlessly integrated, employing image similarity and vectorizer techniques.

Sentiment Analysis - Using NoSQL - Creating Dashboards

Integration of web scraping, data storage in Apache Cassandra, sentiment analysis, and real-time data visualization. It demonstrates the significance of sentiment analysis in news monitoring and high- lights Cassandra’s advantages in efficiently handling and managing for decision making

Analyzing Classical Music

Audio Signal Processing and plotting a spectrogram

Health Insurance Prediction

The aim was to figure out what factors the cost of the premium depends on, and how and to predict the medical insurance costs by using Exploratory Data Analysis and Multiple Linear Regression.

Image-Text Classification Model using CLIP

This is an Image-Text classfication model, built using CLIP, an open-source model by openAI It has a text encoder and an Image Encoder (VIT of ResNet CNN's) that create a joint embedding of the text & image vectors This particular project uses the ViT Image encoder.