Once a business plan and formulation of the problem to be solved in data analysis or data engineering or machine learning solutions, the next phase is about data collection and preparation.
Data processing steps include data collection and intergration, data preprocesssing and data visualization, and feature engineering.
example: how to route customers to agents with the right skills thus reducing call transfers.
How can we predict which skill would solve a customer call..
data collection and intergration ensures the raw data is in once centrally accessible place.
data preprocessing and data visualization involves transforming raw data into an understandable format. this involves data cleaning and exploratory data analysis.
at this stage, we exclude unnecesary labels, entirely inaccurate labels and even combine simmilar labels so as to simplify our model.
data visualization hepps give a quick sense of features and labels summaries. this helps better understand the data and so on.
feature engineering is the process of creating and extracting variables from data
example: we want to base predictions on past data from customer service calls, thus supervised learning/
training our model on historical data that include correct labels or agent skills. then the model can make it own predictions on simmilar data moving forward. the data we need comes from asking questions that help establish our features.
Top comments (0)