Effective Model Update for Adaptive Classification of Text Streams in a Distributed Learning Environment.

In this study, we propose dynamic model update methods for the adaptive classification model of text streams in a distributed learning environment. In particular, we present two model update strategies: (1) the entire model update and (2) the partial model update. The former aims to maximize the model accuracy by periodically rebuilding the model based on the accumulated datasets including recent datasets. Its learning time incrementally increases as the datasets increase, but we alleviate the learning overhead by the distributed learning of the model. The latter fine-tunes the model only with a limited number of recent datasets, noting that the data streams are dependent on a recent event. Therefore, it accelerates the learning speed while maintaining a certain level of accuracy. To verify the proposed update strategies, we extensively apply them to not only fully trainable language models based on CNN, RNN, and Bi-LSTM, but also a pre-trained embedding model based on BERT. Through extensive experiments using two real tweet streaming datasets, we show that the entire model update improves the classification accuracy of the pre-trained offline model; the partial model update also improves it, which shows comparable accuracy with the entire model update, while significantly increasing the learning speed. We also validate the scalability of the proposed distributed learning architecture by showing that the model learning and inference time decrease as the number of worker nodes increases.

Induced pluripotent stem cell models from X-linked adrenoleukodystrophy patients.

OBJECTIVE: Because of a lack of an appropriate animal model system and the inaccessibility of human oligodendrocytes in vivo, X-linked adrenoleukodystrophy (X-ALD)-induced pluripotent stem cells (iPSCs) would provide a unique cellular model for studying etiopathophysiology and development of therapeutics for X-ALD. METHODS: We generated and characterized iPSCs of the 2 major types of X-ALD, childhood cerebral ALD (CCALD) and adrenomyeloneuropathy (AMN), and differentiated them into oligodendrocytes and neurons. We evaluated disease-relevant phenotypes by pharmacological and genetic approaches. RESULTS: We established iPSCs from the patients with CCALD and AMN. Both CCALD and AMN iPSCs normally differentiated into oligodendrocytes, the cell type primarily affected in the X-ALD brain, indicating no developmental defect due to the ABCD1 mutations. Although low in X-ALD iPSCs, very long chain fatty acid (VLCFA) level was significantly increased after oligodendrocyte differentiation. VLCFA accumulation was much higher in CCALD oligodendrocytes than AMN oligodendrocytes but was not significantly different between CCALD and AMN neurons, indicating that the severe clinical manifestations in CCALD might be associated with abnormal VLCFA accumulation in oligodendrocytes. Furthermore, the abnormal accumulation of VLCFA in the X-ALD oligodendrocytes can be reduced by the upregulated ABCD2 gene expression after treatment with lovastatin or 4-phenylbutyrate. INTERPRETATION: X-ALD iPSC model recapitulates the key events of disease development (ie, VLCFA accumulation in oligodendrocytes), provides new clues for better understanding of the disease, and allows for early and accurate diagnosis of the disease subtypes. X-ALD oligodendrocytes can be a useful cell model system to develop new therapeutics for treating X-ALD.