Exploratory data analysis in melamine faced boards

This nugget is about fundamental data analysis for machine learning. A toolbox is part of the nugget, including data visualization, data understanding and data cleaning. A model is only as effective as the capability to obtain the required data and data collection is an essential stage for any successful algorithm performance.

Several examples are shown related to melamine faced boards. Techniques are learned to understand and improve the data you have. Python is used to visualize data. Data should be gathered in one accessible format to make cleaning and understanding easier. Techniques like principal component analysis and Pearson’s correlation threshold are mentioned. Real-life data is imperfect and noisy. Data understanding is essential, and the quality of a model is strongly influenced by the data quality.