Cytogenetics laboratory tests are among the most important procedures for the diagnosis of genetic diseases, especially in the area of hematological malignancies. Manual chromosomal karyotyping methods are time consuming and labor intensive and, hence, expensive. Therefore, to alleviate the process of analysis, several attempts have been made to enhance karyograms. The current chromosomal image enhancement is based on classical image processing. This approach has its limitations, one of which is that it has a mandatory application to all chromosomes, where customized application to each chromosome is ideal. Moreover, each chromosome needs a different level of enhancement, depending on whether a given area is from the chromosome itself or it is just an artifact from staining. The analysis of poor-quality karyograms, which is a difficulty faced often in preparations from cancer samples, is time consuming and might result in missing the abnormality or difficulty in reporting the exact breakpoint within the chromosome. We developed ChromoEnhancer, a novel artificial-intelligence-based method to enhance neoplastic karyogram images. The method is based on Generative Adversarial Networks (GANs) with a data-centric approach. GANs are known for the conversion of one image domain to another. We used GANs to convert poor-quality karyograms into good-quality images. Our method of karyogram enhancement led to robust routine cytogenetic analysis and, therefore, to accurate detection of cryptic chromosomal abnormalities. To evaluate ChromoEnahancer, we randomly assigned a subset of the enhanced images and their corresponding original (unenhanced) images to two independent cytogeneticists to measure the karyogram quality and the elapsed time to complete the analysis, using four rating criteria, each scaled from 1 to 5. Furthermore, we compared the enhanced images with our method to the original ones, using quantitative measures (PSNR and SSIM metrics).

Deep learning natural language processing models often use vector word embeddings, such as word2vec or GloVe, to represent words. A discrete sequence of words can be much more easily integrated with downstream neural layers if it is represented as a sequence of continuous vectors. Also, semantic relationships between words, learned from a text corpus, can be encoded in the relative configurations of the embedding vectors. However, storing and accessing embedding vectors for all words in a dictionary requires large amount of space, and may stain systems with limited GPU memory. Here, we used approaches inspired by quantum computing to propose two related methods, word2ket and word2ketXS, for storing word embedding matrix during training and inference in a highly efficient way. Our approach achieves a hundred-fold or more reduction in the space required to store the embeddings with almost no relative drop in accuracy in practical natural language processing tasks.

Quantum machine learning methods have the potential to facilitate learning using extremely large datasets. While the availability of data for training machine learning models is steadily increasing, oftentimes it is much easier to collect feature vectors to obtain the corresponding labels. One of the approaches for addressing this issue is to use semi-supervised learning, which leverages not only the labeled samples, but also unlabeled feature vectors. Here, we present a quantum machine learning algorithm for training semi-supervised kernel support vector machines. The algorithm uses recent advances in quantum sample-based Hamiltonian simulation to extend the existing quantum LS-SVM algorithm to handle the semisupervised term in the loss. Through a theoretical study of the algorithm's computational complexity, we show that it maintains the same speedup as the fully-supervised quantum LS-SVM.

In this paper, we investigate a commonsense inference task that unifies natural language understanding and commonsense reasoning. We describe our attempt at SemEval-2020 Task 4 competition: Commonsense Validation and Explanation (ComVE) challenge. We discuss several state-of-the-art deep learning architectures for this challenge. Our system uses prepared labeled textual datasets that were manually curated for three different natural language inference subtasks. The goal of the first subtask is to test whether a model can distinguish between natural language statements that make sense and those that do not make sense. We compare the performance of several language models and fine-tuned classifiers. Then, we propose a method inspired by question/answering tasks to treat a classification problem as a multiple choice question task to boost the performance of our experimental results (96.06%), which is significantly better than the baseline. For the second subtask, which is to select the reason why a statement does not make sense, we stand within the first six teams (93.7%) among 27 participants with very competitive results. Our result for last subtask of generating reason against the nonsense statement shows many potentials for future researches as we applied the most powerful generative model of language (GPT-2) with 6.1732 BLEU score among first four teams 1 .

No abstract