Ankit Upadhyay

🕵️‍♀️ I am Ankit Upadhyay, Third Year Undergraduate at National Institute of Technology, Allahabad, India.

🛠️ I am passionate about Machine Learning and have gained experience in Time Series Analysis, Natural Language Processing and Reinforcement Learning

🌍 As an open-source contributor, I’m getting involved in the community to help improve tools and libraries.

My Experiences

Open Source Contribution 🤝

I’m currently working under the mentorship program of sktime, a framework for time-series analysis. It’s been a rewarding experience adding features and learning from the open-source community while enhancing my skills in time-series forecasting and model development.

Implementation 💡

I’m very interested in having a in depth understanding of different ML algorithms. To explore this, I’ve implemented some key research papers.

  • I have implemented Proximity Policy Optimization(PPO) Algorithm for Cartpole Problem. Check it out here ppo-cart
  • I’ve worked on implementing KMeans Clustering for image compression

    Image compression image

  • I have Implemented DQN Algo from Scratch and a wrapper for Chess environment.

    Image compression image

    Check it out at My Implementations

    FAIR AI

    🎯 I competed in the Enigma ML competition, focusing on building a model that balances accuracy and fairness, ensuring precision without bias.

Through this, I also learned how to use Optuna for hyperparameter tuning and applied in-depth data analysis to enhance model performance.

Graph image