Machine Learning uses statistical techniques to give computer systems the ability to “learn” with data without being explicitly programmed.

Steps in Machine Learning

In general, there are 5 basic steps involved in ML

  1. Data Collection: Be it the raw data from excel, access, text files any source, this step forms the foundation of future learning.
  2. Data Preparation: Quality of data plays a critical role in ML. One should spend more time on determining the quality of data and fixing any nuances observed (like missing data points). Data is cleaned and prepared for learning during this step. The cleaned data is usually split into 2 sets in the ratio of 80:20.
  3. Model Training: This step involves choosing the appropriate algorithm and the data representation in the form of a model. Training Dataset — 80% of the cleaned data is used for training the model.
  4. Model Evaluation: This step is used to evaluate the model and the chosen algorithm. The accuracy of a model can be evaluated by analyzing its performance on data that was not used at all during model training. To test the accuracy of the Model, the remaining 20% of the cleaned data (Test Dataset) is used.
  5. Improving Performance: Based on the Model evaluation, this step might involve choosing a different model altogether or introducing more variables to improve the efficiency of the model.