Nasa Nearest Earth Objects

Machine Learning

Once the data had been prepped for predictive modeling, I was able to build models using 4 different machine learning techniques.

Insights from Visualizations

With any data project, visualizations play a key role in deriving insight. I found the use of pair plots, bar charts, and histograms to be useful for identifying trends in the data and also for identifying desirable parameter configurations for certain models.

Data Exploration

The objective of this project was to use the dataset to train a model that could accurately predict if an asteroid would be considered a "near-earth" object. Before building the model I performed an exploratory data analysis. The pair plot proved to be a useful tool for observing the divide in the data points.