machinelearningprojects

I built a classification model using machine learning tools in Databricks. The aim is to classify whether there is a fault with a manufacturing machine based on the readings from the vibration sensors. Two other binary classification algorithms, viz, Decision Tree and Random Forest, were employed to build two different models, and their respective performances are compared using the customer dataset sourced from a Telecom company in Iran. This task aims to build two different binary-classification models and recommend a more efficient model for identifying customers who are likely to churn, that is, customers that are likely to stop patronising the company. As part of the industry standard recommendation, the Cross Industry Standard Process for Data Mining (CRISP-DM) methodology was strictly adhered to in this task. Also, two different clustering algorithms, namely, DB-SCAN and K-Means algorithms, were applied to build two clustering models that can estimate the obesity levels of individuals from Mexico, Peru and Colombia based on their eating habits and physical conditions. The clustering models aim to identify the clusters in the dataset, which should likely translate into the existing, different estimates of obesity levels across the set of individuals and analyse the models’ respective performances.