Abstract:Today, modern databases with "Big Dimensionality" are experiencing a growing trend. Existing approaches that require the calculations of pairwise feature correlations in their algorithmic designs have scored miserably on such databases, since computing the full correlation matrix (i.e., square of dimensionality in size) is computationally very intensive (i.e., million features would translate to trillion correlations). This poses a notable challenge that has received much lesser attention in the field of machine learning and data mining research. Thus, this paper presents a study to fill in this gap. Our findings on several established databases with big dimensionality across a wide spectrum of domains have indicated that an extremely small portion of the feature pairs contributes significantly to the underlying interactions and there exists feature groups that are highly correlated. Inspired by the intriguing observations, we introduce a novel learning approach that exploits the presence of sparse correlations for the efficient identifications of informative and correlated feature groups from big dimensional data that translates to a reduction in complexity from O(m2n) to O(mlogm + Ka mn), where Ka << min(m,n) generally holds. In particular, our proposed approach considers an explicit incorporation of linear and nonlinear correlation measures as constraints in the learning model. An efficient embedded feature selection strategy, designed to filter out the large number of non-contributing correlations that could otherwise confuse the classifier while identifying the correlated and informative feature groups, forms one of the highlights of our approach. We also demonstrated the proposed method on one-class learning, where notable speedup can be observed when solving one-class problem on big dimensional data. Further, to identify robust informative features with minimal sampling bias, our feature selection strategy embeds the V-fold cross validation in the learning model, so as to seek for features that exhibit stable or consistent performance accuracy on multiple data folds. Extensive empirical studies on both synthetic and several real-world datasets comprising up to 30 million dimensions are subsequently conducted to assess and showcase the efficacy of the proposed approach.

Regression Based on Neural Incremental Attribute Learning with Correlation-Based Feature Ordering

Ordered Incremental Attribute Learning based on mRMR and Neural Networks

Neural Incremental Attribute Learning in Groups.

Feature Ordering for Neural Incremental Attribute Learning Based on Fisher's Linear Discriminant

OPTIMIZED NEURAL INCREMENTAL ATTRIBUTE LEARNING FOR CLASSIFICATION BASED ON STATISTICAL DISCRIMINABILITY

Sparse Feature Preservation for Relative Attribute Learning.

Linear Feature Sensibility for Output Partitioning in Ordered Neural Incremental Attribute Learning.

Feature selection in growing dimensional space for classification based on neural incremental attribute learning

Click-Through Rate Prediction Algorithm Based on Modeling of Implicit High-Order Feature Importance

Learning Robust Features with Incremental Auto-Encoders

Feature Discriminability for Pattern Classification Based on Neural Incremental Attribute Learning

Statistical Discriminability Estimation for Pattern Classification Based on Neural Incremental Attribute Learning

Neural Incremental Attribute Learning Based on Principal Component Analysis

Evolving Linear Discriminant in a Continuously Growing Dimensional Space for Incremental Attribute Learning.

Pattern Classification with Ordered Features Using Mrmr and Neural Networks

Output Effect Evaluation Based on Input Features in Neural Incremental Attribute Learning for Better Classification Performance

Integrated Feature Preprocessing for Classification Based on Neural Incremental Attribute Learning.

Making Trillion Correlations Feasible in Feature Grouping and Selection

Feature Estimations Based Correlation Distillation for Incremental Image Retrieval

Cost-sensitive Regression Learning on Small Dataset Through Intra-Cluster Product Favoured Feature Selection

ICFD: an Incremental Learning Method Based on Data Feature Distribution.