xaviercarrera / platzi-master-ml-exercises Goto Github PK

Here you can find some of the challenges for Platzi Master -- an education acceleration program to form the future tech leaders in LATAM.

The main areas concerning these challenges are:

• Business Analytics
• Applied Statistics
• Data Science and Data Engineering
• Machine Learning and Artificial Intelligence

1. Mall Clients Data

A mall customer dataset contains the details about the people who visit the mall. The dataset has age, customer ID, gender, annual income, and expense score. The objective is to gain an insight into data and divide customers into different groups based on their behavior.

2. Boston Housing Analysis

We predicted housing prices using linear regression, PCA and and random forests. These algorithms are used to predict unknown values when the data have some linear relationship between input and output variables. The final outcome is a R² score of 88%.

3. Uber Pickups Data Visualization

The dataset has information on 4.5 million Uber pickups in New York City from April 2014 to September 2014. The idea of the project is to analyze customer journey data and visualize the data to find information that can help improve the business.

4. Amazon Product Recommendation System

We take over 500k instruments reviews from Amazon to build a recommendation algorithm. In this sense, we use the Surprise library to build an algorithm based on KNN and collaborative filtering.

5. Breast Histopathology Data Analysis

This dataset contains 2.77,524 50x50 size images drawn from 162 montage slide images of breast cancer specimens scanned at 40x. There are 1,98,738 negative tests and 78,786 positive tests with IDC. The challenge is to build a model that can classify breast cancer, building an image classification model with convolutional neural networks.