Info originates from different sources, paying homage to sensor measurements, events, textual content material, photographs, and flicks, with the Net of Points (IoT) producing regular streams of information.
A variety of this information is unstructured, like photographs consisting of pixel RGB values, texts composed of phrases and characters, and clickstreams of particular person actions. A key downside in data science is altering this raw, unstructured data into actionable and structured form.
Structured data is offered in two main kinds: numeric and categorical.
Numeric data could also be regular, like temperature or distance, and discrete, identical to the number of autos in a car parking zone.
Categorical data has a set set of values, paying homage to automotive fashions (sedan, SUV, hatchback) or metropolis names (New York, Los Angeles). A selected case of categorical data is binary data, which has two values, paying homage to 0/1 or on/off. Ordinal data, one different type of categorical data, has ordered lessons, like tutorial ranges (highschool, bachelor’s, grasp’s, doctorate).
References
- Wise Statistics for Info Scientists: 50+ Vital Concepts using R and Python
