Department:  Petroleum and Mining  Engineering

The workshop in brief:

The aim of this workshop is to present various methods for handling missing data within a dataset. Several approaches will be introduced and illustrated with practical examples.

Participants: 16 staff participants.

Place: Tishk International University, Hall 302, Ed. Building, TIU

Session topics: 

   Session I: Introduction, Data analysis , position of missing data within data analysis, preprocesses within data cleaning , Why data is missing, importance of handling missing values, symbols used to reflect  cells with missing data,  harmful effects of missing data if it is not handled properly, challenges Posed by missing Values, mechanisms of missing data ( MCAR, MAR, MNAR), tree diagram for techniques for handling missing data.

  Session II: Factors affecting selection of the proper technique for handling missing data, application of techniques used for  handling missing data with solved examples, mean, median, and mode imputations, listwise and pairwise deletion, deterministic and stochastic regression imputation, KNN method, Hot and Cold Deck methods, stochastic multiple imputation, maximum likelihood  estimation, interpolation technique, LOCF, NOCB,  Regression decision tree. Advantages and disadvantages of different techniques. Discussion of the addressed questions.

Professor Dr. Tariq Hamakareem

• The staff members attended the sessions made benefits from the materials presented as was understood from the discussions.