Skip to main content
Institute of Advanced Technological and Commercial Studies

Glossary

Dataset: the raw material of a model

A dataset gathers the observations an analysis or a model rests on. Its preparation decides the reliability of the result long before the algorithm is chosen. Assembling, harmonising, checking representativeness: most of the work happens there.

What a dataset must carry

A dataset gathers observations described by variables. Preparation takes up most of the work: bringing sources together, harmonising formats, handling missing values, checking the sample genuinely represents the target population, then setting aside a share for evaluation. The data and artificial intelligence field holds 67 modules in the catalogue.

  • Observations, variables and a clearly stated target population
  • The evaluation share is set aside before any training begins
  • 67 modules in the data and artificial intelligence field

From raw data to a ready dataset

The Data Science programme has candidates assemble a complete dataset from heterogeneous sources, through to the training and evaluation split. The Artificial Intelligence programme works on representativeness: what the dataset contains determines what the model will be able to handle once in service.

  1. 01

    1 — Bring the sources together

    Internal databases, operational files, field readings: each source arrives in its own format.

  2. 02

    2 — Harmonise

    Units, dates, identifiers and labels are brought to a single documented form.

  3. 03

    3 — Check representativeness

    The sample is compared with the target population: that check is what makes the result transferable.

  4. 04

    4 — Set aside the evaluation

    A share is set aside before any training and serves only to measure the final model.

Programmes that have you prepare one

2 catalogue programmes put “Dataset” to work: 240 credits and 2400 taught hours in total. The official rule holds throughout — one credit stands for 25 hours of work, 10 of them taught. Every line below is recomputed from the programme page: level, duration, credits, taught volume and fees in Guinean francs appear exactly as filed in the official catalogue.

ProgrammeLevelDurationVolumeFees
Data Science programme — Bac+5 levelMaster's degree2 years120 credits · 1200 taught hours7 200 000 GNF per year
Artificial Intelligence programme — Bac+5 levelMaster's degree2 years120 credits · 1200 taught hours7 800 000 GNF per year
Dataset: the raw material of a model — Glossary | IHETC — IHETC