Plaban’s Portfolio

My Data Science Portfolio!

Project 1: MedicareAI

Implemented a flask web application with machine learning for instant medical checkups for the patients, which can help doctors/healthcare providers. By analyzing the symptoms Medicare AI system determines the severity of the disease in near real-time and guides the patient with – disease definition, precaution, and recommendations.

Project 2: Medicine Recommendation using Sentiment Analysis

Performed data analysis for the reviews collected from the patients for the Medicine/Drug name recommendation. In the data exploration part, we used visualization and statistical analysis techniques to understand data. Through this process, we determined the topic to preprocess the data which will fit the objective. We have then created different variables to fit the machine learning model. For the modelling part, we implemented the emotional sentiment analysis using the word dictionary by Harvard, n-gram with deep learning etc.

Project 3: Cash Machine Trading Simulator

An web application to simulate stock trading Simulator. Fetching stock data Using yahoo finance API . Implemented the “Simple Moving Average” algorithm to understand the stock performance over the years and predict buy/sell signals accordingly. Used flask packages to develop a web application to view the simulation results according to user preferences.

Project 4: Ingest transform and visualize customer product reviews dataset using aws sagemaker

Ingest and transform the customer product reviews dataset using AWS data stack services such as AWS Glue and Amazon Athena for ingesting and querying the dataset. And using AWS Data Wrangler to analyze the dataset and plot some visuals extracting insights.

Project 5: Detect-data-bias-with-Amazon-SageMaker-Clarify

Bias can be present in our data before any model training occurs. Inspecting the dataset for bias can help detect collection gaps, inform our feature engineering, and understand societal biases the dataset may reflect.We will analyze bias on the dataset, generate and analyze bias report, and prepare the dataset for the model training.

Project 6: Train a model with Amazon SageMaker Autopilot

Using Amazon Sagemaker Autopilot to train a BERT-based natural language processing (NLP) model. The model will analyze customer feedback and classify the messages into positive (1), neutral (0) and negative (-1) sentiment.

Project 7: Applied_ML_with_5_different_algorithms

Applied 5 different algorithm on Titanic Dataset to build model to predict which people would survive based on some given attributes. To solve this classification problem used the following angorithms and then compared model results and final model selection was done. Logistic Regression Algorithm, Support Vector Machines Algorithm, Multilayer Perceptron Algorithm, Random Forest, GradientBoosting

Project 8: Stock Price Prediction using LSTM

Predicting stock pricing using Long short-term memory.

Project 9: Detecting Credit Card Fraud

Detecting Credit Card Fraud using different machine learning and deep learning algorithms

Project 10: Extreme-Gradient-Boosting-with-XGBoost

Topics covered: - · Using XGBoost for classification tasks · Using XGBoost for Regression tasks · Fine-tuning XGBoost’s most important hyperparameters · Incorporating XGBoost in pipelines