A predictive model is an effective tool supporting business decisions. The creation of a model is a process of searching for patterns using the machine learning algorithms in the existing set of data (model training) to predict the most likely future scenarios (model application).
A classification model allows you to split a large array of data into groups (classes) that are characterized by a similar set of characteristics, using explicit and implicit regularities. The use of such a model allows, for example, to define groups of users who are characterized by a certain behavior and make special suggestions to them.
To ensure the quality of the model it is necessary to select and prepare input data carefully. There may also be a need for external data. The result of my work will be:
1) the model itself in the format determined by the model chosen (for example, in the case of linear models, this will be the vector of weights of each of the factors. The weight of the factor determines the effect of its influence on the result)
2) a report on the validity (prediction accuracy) of the model
3) a file with the results of applying the model to the data set being examined