E-Locus - Institutional Repository of the University of Crete

Home Collections School/Department School of Sciences and Engineering Department of Computer Science Post-graduate theses

Post-graduate theses

Current Record: 30 of 840

[Add to Basket]

Identifier

000460153

Title

Do we need sophisticated imputation in AutoML?

Alternative Title

Χρειαζόμαστε σύνθετη συμπλήρωση ελλειπουσών τιμών στην Αυτοματοποιημένη Μηχανική Μάθηση;

Author

Πατεράκης, Γιώργος Μ.

Thesis advisor

Τσαμαρδινός, Ιωάννης

Reviewer

Χριστοφίδης, Βασίλης
Τριανταφύλλου, Σοφία

Abstract

Numerous real-world data contain missing values, while in contrast, most Machine Learning (ML) algorithms assume complete datasets. For this reason, several imputation algorithms have been proposed to predict and fill in the missing values. Given the advances in predictive modeling algorithms tuned in an AutoML setting, a question that naturally arises is to what extent sophisticated imputation algorithms (e.g., Neural Network based) are really needed, or we can obtain a decent performance using simple methods like Mean/Mode (MM). In this paper, we experimentally compare 6 state-of-the-art representatives of different imputation algorithmic families from an AutoML predictive modeling perspective, including a feature selection step and combined algorithm and hyper-parameter selection. Experiments ran on 25 real-world incomplete datasets with missing values and 10 complete datasets in which synthetic missing values are introduced according to different missingness mechanisms, at varying missing frequencies. The main conclusion drawn from our experiments is that the best method on average is the Denoise AutoEncoder (DAE) on real-world datasets and the MissForest (MF) in simulated datasets, followed closely by Mean/Mode. In addition, binary indicator (BI) variables encoding missingness patterns actually improve predictive performance, on average. Last but not least, although there are cases where Neural-Networkbased imputation significantly improves predictive performance, this comes at a great computational cost and requires measuring all feature values to impute new samples.

Language

English

Subject

Machine learning

Missing data

Ελλειπούσες τιμές

Μηχανική μάθηση

Issue date

2023-12-01

Collection

School/Department--School of Sciences and Engineering--Department of Computer Science--Post-graduate theses

Type of Work--Post-graduate theses