Kunal Biswas

Transforming Data into Intelligence with Python-Powered AI/ML Solutions. Leveraging advanced algorithms and deep learning techniques to drive innovation and optimize performance. Proven expertise in developing scalable, efficient machine learning models and deploying them in production environments. Dedicated to enhancing data-driven decision-making and delivering impactful results.

NLP(ude.my/UC-1a54c6ac-44df-4fb7-a1b1-8ed5d541fde4)

• Medical Aid Reimbursement(Fraud Detection Analysis)

As per my project in current company is related to healthcare in which an deep learning model(ANN) is used to predict whether patient is eligible for medicare or medicaid and if medicaid upto what extent.

• Hate Speech Detection(for learning purpose)

I have used twitter dataset from kaggle using NLP(NLTK and Spacy) for preprocessing(TF-IDF for word vectorization) the data. Then used Deep Learning Model (RNN)

• Environmental sound detection( For learning purpose)

Used python Package LIBEROSA for feature extraction of audio dataset and used Machine Learning model(Random Forest) using ML framework from Scikit Learn

• Stock Market Prediction(Predictive Analysis)

Used deep learning model(ANN) using DL framework tensorflow and keras.

• Cat and Dog Classification( For learning purpose)

Using both keras.preprocessing.image(with CNN) and skimage(with SVM)

• Speech to Text and Text to Speech(For learning Purpose)

Using Google Text to Speech(gTTS)

• Medical Aid Reimbursement (Fraud Detection Analysis)In my current role, I developed a deep learning model (ANN) to predict patient eligibility for Medicare or Medicaid and determine the extent of Medicaid coverage. This project focuses on fraud detection in the healthcare sector.
• Hate Speech Detection (Learning Project)Utilized a Twitter dataset from Kaggle, leveraging NLP techniques (NLTK and SpaCy) for preprocessing and TF-IDF for word vectorization. Implemented a Deep Learning model (RNN) to detect hate speech.
• Environmental Sound Detection (Learning Project)Employed the LIBROSA package for feature extraction from audio datasets and developed a Machine Learning model (Random Forest) using Scikit-Learn for environmental sound classification.
• Stock Market Prediction (Predictive Analysis)Built and deployed a deep learning model (ANN) using TensorFlow and Keras to predict stock market trends and inform trading strategies.
• Cat and Dog Classification (Learning Project)Developed image classification models using both Keras (with CNN) and Scikit-Image (with SVM) to distinguish between cat and dog images.
• Speech to Text and Text to Speech (Learning Project)Implemented Google Text-to-Speech (gTTS) for converting speech to text and vice versa, enhancing user accessibility and interaction