Sue Sin Chong

Sue Sin Chong

Is it possible to train several classifiers to perform meaningful crowd-sourcing to produce a better prediction label set without ground-truth annotation? This paper will modify the contrastive learning objectives to automatically train a self-complementing ensemble to produce a state-of-the-art prediction on the CIFAR10 and CIFAR100-20 tasks. This paper will present a straightforward method to modify a single unsupervised classification pipeline to automatically generate an ensemble of neural networks with varied decision boundaries to learn a more extensive feature set of classes. Loss Function Entropy Regularization (LFER) are regularization terms to be added to the pre-training and contrastive learning loss functions. LFER is a gear to modify the entropy state of the output space of unsupervised learning, thereby diversifying the latent representation of decision boundaries of neural networks. Ensemble trained with LFER has higher successful prediction accuracy for samples near decision boundaries. LFER is an adequate gear to perturb decision boundaries and has produced classifiers that beat state-of-the-art at the contrastive learning stage. Experiments show that LFER can produce an ensemble with accuracy comparable to the state-of-the-art yet have varied latent decision boundaries. It allows us to perform meaningful verification for samples near decision boundaries, encouraging the correct classification of near-boundary samples. By compounding the probability of correct prediction of a single sample amongst an ensemble of neural network trained, our method can improve upon a single classifier by denoising and affirming correct feature mappings.

Sue Sin Chong

Most of the current inference techniques rely upon Bayesian inference on Probabilistic Graphical Models of observations and do predictions and classification on observations. However, there is very little literature on the mining of relationships between observations and building models of the relationship between sets of observations or the generating context of the observations. Moreover, event understanding of machines with observation inputs needs to understand the relationship between observations. Thus there is a crucial need to build models and develop effective data structures to accumulate and organize relationships between observations. Given a PGM model, this paper attempts to fit a permutation of states to a sequence of observation tokens (The Travelling Photographer Problem). We have devised a machine learning inspired architecture for randomized approximation of state permutation, facilitating parallelization of heuristic search of permutations. As a result, our algorithm can solve The Travelling Photographer Problem with minimal error. Furthermore, we demonstrate that by mimicking machine learning components such as normalization, dropout, and lambda layer with a randomized algorithm, we can devise an architecture that solves TPP, a permutation NP-Hard problem. Other than TPP, we can also provide a 2-Local improvement heuristic for the Travelling Salesman Problem (TSP) with similar ideas.