ITPEC IP October 2024 Question 94

Machine learning, in addition to statistical analysis, is effective for forecasting — the only statement that does not wrongly restrict what big data analysis can use.

Big data is characterized by huge volume, a wide variety of forms and high speed of generation. Statistical analysis finds relationships and trends in it, and machine learning learns patterns from large data sets to predict unknown cases; combining the two is a normal approach to forecasting.

Answer (a)

Why not others:

• •(b) — video, images and audio are part of big data's variety and are analyzed with image and speech recognition
• •(c) — human actions are also analyzed from machine-generated data such as sensor, location, IC card and purchase logs, not only from what people post or search
• •(d) — data not created for analysis, such as blog entries, is often analyzed (for example by text mining for customer opinion); what counts is the purpose of the analysis, not why the data was written

Key rule: Big data spans structured and unstructured data from any source; a statement that rules out a data type or source outright is usually wrong.

System Strategy · AI-written, may contain errors