A two-stage predictive machine learning engine that forecasts the on-time performance of flights for 15 different airports in the USA based on data collected in 2016 and 2017.

The aim to decrease the maintenance cost of generators used in wind energy production machinery. This is achieved by building various classification models, accounting for class imbalance, and tuning on a user defined cost metric (function of true positives, false positives and false negatives predicted) & productionising the model using pipelines.

Machine Learning model API deployed with GCP & DVC. Predicts a breast-cancer diagnostic.

Credit card fraud detection using machine learning techniques

Credit Card Fraud Detection using Python and Machine Learning.

Code to detect credit card fraud detecton

A fraud detection project that processes user or credit card data using machine learning and deep learning algorithms.

Support Vector Machine Classification model is applied on bank dataset containing 41188 rows and 21 columns. The data is related with direct marketing campaigns of a Portuguese banking institution. The marketing campaigns were based on phone calls. Often, more than one contact to the same client was required, in order to assess if the product (b…