shasha2408 / datasets Goto Github PK

Multi variate time series analysis - outlier / anomaly identification.

There are 509k samples with 11 features.

Each instance / row is one moment in time.

0.09% of rows are identified as events, basically rare, outliers / anomalies

Task is to identify these events in this time series - time series classification

As dataset is unbalanced, the metric to use is F1 score(not accuracy, nor ROC AUC)

Dataset Description: The analysis of data concerning the wafer production process with the goal of determining possible causes for errors, resulting in lots of faulty wafers.

Train.csv - 1763 rows x 1559 columns

Test.csv - 756 rows x 1558 columns

Attribute Description:

Feature1 - Feature1558 - Represents the various attributes that were collected from the manufacturing machine

Class - (0 or 1) - Represents Good/Anomalous class labels for the products

optimizing Area under the curve(AUC) to generalize well on unseen data