Awritrojit Banerjee
Student
Erlangen
Summary
Motivated student currently working towards a masters' degree in Artificial Intelligence. Adept at prepping research, software architecture design, pipelining in ML projects and developing extensive and custom designed AI algorithms to solve specific tasks. Extensive background in investigating Pattern Recognition research in Swarm Intelligence, Statistical Machine Learning, Bayesian Machine Learning and Deep Learning especially in CNNs and LSTMs.
Overview
3
3
years of professional experience
6
6
Certifications
3
3
Languages
Work History
Research Assistant for Urban Physics Modelling
Fraunhofer Institute For Building Physics
05.2023 - 01.2024
- Worked both independently and collaboratively in fast-paced experimentation environment.
- Performed statistical, qualitative, and quantitative analysis.
- Developed state-of-the-art AI/ML models to solve the super-resolution/statistical downscaling problem pertaining to urban physics spatiotemporal data.
- Successfully conducted regression analysis, AI pipeline design, and delivered explainable presentations based on the results obtained from complex ML algorithms.
Analytics Intern
Centre For Development Of Advanced Computing
06.2022 - 10.2022
- Successfully completed an image recognition task - developed and trained a simple CNN model to predict the shelf life and corresponding biochemical profile of plantains given their current image, and this model was especially developed and optimized on a small dataset of only 86 images.
- Introduced and developed the pipleine for IMU signal processing using 1D Convolutions in Tensorflow.
- Developed the datapipeline to read raw signal data and convert it into a format suitable for interfacing with the model pipeline.
Research Intern
Jadavpur University
04.2021 - 05.2022
- Studied biological cybernetics focusing on swarm intelligence. Used swarm based algorithms for non-linear feature-space optimization (feature/dimensionality reduction).
- Developed a novel classification algorithm that can detect non-contextual datapoints with an accuracy above 95%. The classifier, tested on the Iris and Wisconsin Breast Cancer dataset gave a classification accuracy of more than 92% in each case. A publication for this project is underway.
Education
Master of Science - Artificial Intelligence
Friedrich Alexander University Erlangen-Nuremberg
Erlangen, Bavaria 04.2001 -
Bachelor of Technology - Information Technology
St. Thomas' College of Engineering & Technology
Kolkata, India 04.2001 -
Skills
Statistical modelling: ScikitLearn, Numpy, Pandas
Deep Learning: CNN, LSTM, TCN, ConvLSTM, Tansformers, BERT, Autoencoders, MLP, NLP, Tensorflow, PyTorch
Signal processing: Neuromuscular signal processing using EMG
Languages: Python, Matlab, C, C
Optimization: Swarm based methods, ILP, Gurobi
ML model end-to-end architecture design
Data and model pipelining
Document Management
Certification
Machine Learning Specialization
Timeline
Research Assistant for Urban Physics Modelling
Fraunhofer Institute For Building Physics
05.2023 - 01.2024
Analytics Intern
Centre For Development Of Advanced Computing
06.2022 - 10.2022
Research Intern
Jadavpur University
04.2021 - 05.2022
Master of Science - Artificial Intelligence
Friedrich Alexander University Erlangen-Nuremberg
04.2001 -
Bachelor of Technology - Information Technology
St. Thomas' College of Engineering & Technology
04.2001 -