I’m currently focusing on building products powered by AI and large language models (LLMs).

Revanth Vermareddy  • Peerlist

The S&P 500 is widely considered as a standard when analysing the US stock markets. The project detects sudden fluctuations in a time series data like the index data over and above a certain threshold. Used keras api backed by tensorflow and built an LSTM auto encoder to predict the anomaly behavior. The model accuracy is about 98%.

Anomaly Detection in S&P500 Index

In this project, we analysed the EEG signals (brain signals) of the patients suffering from Epilepsy (a neurological disorder). Based on the energy patterns, levels of intensity and unusual patterns which help identify a normal person from a person with seizures. We developed an ML classification model to predict the impending seizure attack beforehand and built an altering system in case of emergency. Benchmarked the model against various state of art classifiers and the promising results were obtained from Random Forest Classifier. The model is able to predict with an accuracy of 99%.

EEG Signal Classification For Epileptic Seizu

This project is a movie recommendation system which uses machine learning to recommend the user a movie to watch. Followed a two way approach to handle recommendations for a new year and an existing user. Recommended is made by analyzing a user previous ratings to better understand one's taste and preferences. For a new user it recommends the trending and high rated pick among the existing users.