Welcome to my GitHub profile! I'm a passionate data scientist who can leverage data to drive actionable insights. With a strong background in business analytics and an MBA degree in Business Analytics and Finance, I bring a unique blend of analytical skills and business acumen to my data science projects. I am a Btech. graduate in ECE who used to work previously as a full stack developer.
- Python: I have extensive experience working with Python for data analysis, machine learning, and building data-driven solutions.
- Natural Language Processing (NLP) Chatbots: I specialize in developing intelligent chatbot systems using the Rasa Framework, enabling conversational AI solutions. Also used other chatbot frameworks such as Dialogue Flow and basic understanding of language transformer models such as BERT.
- Finance Domain: My expertise lies in applying data science techniques to solve challenges in the finance industry, such as stock clustering and recommendations, credit default forecasting, risk analysis, and portfolio optimization.
- Business Analytics: With an MBA degree in business analytics, I possess a deep understanding of using data to drive strategic decision-making and solve complex business problems.
- Languages – Python , Core Java, R, PL/SQL
- Framework/Lib – NumPy, Pandas, Scikit-learn ,Tensorflow, Keras, Matplotlib, Seaborn, Plotly
- ML – Regression, Classification, Clustering, Dimensionality Reduction,Ensemble Technique, Feature Selection
- NLP – NLP Tasks(Word Embeddings : NLTK,TFIDF, Word2Vec, Fasttext , Spacy , Gensim , DistilBERT,Classification ,NER,Sentiment Analysis), Chatbot Frameworks( RASA, Dialogflow), Transformers
- DL – CNN, ANN
- IDE – Pycharm, Jupyter Notebooks, Google Colab,Eclipse,R Studio
- Deployment- Streamlit, Flask APIs, Fast API, Docker basics, MongoDB, Gitlab, Git
- Real-time Tweet Sentiment Analysis and Wordclouds Dashboard: This is a repository of a real time tweet sentiment analysis app built using streamlit platform and using Hugging Face Bertweet Model as core classifier model.
- Bankruptcy Prediction in Highly Imbalanced Dataset Using Instance Hardness Threshold (IHT) Undersampling: Bankruptcy Prediction - By integrating Instance Hardness(IHT) based Under sampling and Supervised Learning Methods in highly Imbalanced dataset and SHAP Interpretation of Model
- Hierarchical Risk PArity Based Portfolio Optimization Comparison: In this project we explore the method of HRP for porfolio composition which has well diversified risk. And we compare it's performance against conventional portfolio optimisation technique such as Markowitz Mean-Variance(MVP) Portfolio.
- Probabilistic Neural Network(PNN) for Classification: In this notebook understanding PNN and its related concepts . Concepts of Parzen Window or KDE(kernel density estimate) .Kernel functions as non-parametric method to ascertain data distribution through an example. Implementation of PNN using python for classification tasks.
I believe in continuous learning to stay ahead in the rapidly evolving field of data science. Currently, I'm focusing on:
- Advanced NLP Techniques: Exploring advanced NLP algorithms and models to enhance the capabilities of chatbot systems like LLMs and Generative AI Chat GPT.
- Deep Learning for Finance: Delving into deep learning techniques tailored for finance applications, such as time series analysis and neural networks.
- Meta-Heuristics Techniques: Exploring and reading learning on meta heuristic techniques such as Ant Colony Optimisation, Particle Swarm Optimisation , Tabu Search etc.
- Cloud Computing: Expanding my knowledge of cloud platforms and services to leverage scalable and cost-effective solutions for data science projects.
I'm always open to collaborating on exciting data science projects that align with my interests. If you're working on a project at the intersection of finance, NLP, and data analytics, I would love to discuss how we can work together.
- LinkedIn: LinkedIn Profile URL
- Kaggle Website: Kaggle Profile
- Email: [[email protected]]
Let's connect and explore the fascinating world of data science and analytics together!