Stack LSTM-Based User Identification Using Smart Shoes with Accelerometer Data

In this study, we propose a long short-term memory (LSTM)-based user identification method using accelerometer data from smart shoes. In general, for the user identification with human walking data, we require a pre-processing stage in order to divide human walking data into individual steps. Next, user identification can be made with divided step data. In these approaches, when there exist partial data that cannot complete a single step, it is difficult to apply those data to the classification. Considering these facts, in this study, we present a stack LSTM-based user identification method for smart-shoes data. Rather than using a complicated analysis method, we designed an LSTM network for user identification with accelerometer data of smart shoes. In order to learn partial data, the LSTM network was trained using walking data with random sizes and random locations. Then, the identification can be made without any additional analysis such as step division. In the experiments, user walking data with 10 m were used. The experimental results show that the average recognition rate was about 93.41%, 97.19%, and 98.26% by using walking data of 2.6, 3.9, and 5.2 s, respectively. With the experimental results, we show that the proposed method can classify users effectively.

Evaluation validation of a qPCR curve analysis method and conventional approaches

Background Reverse Transcription quantitative polymerase chain reaction (RT-qPCR) is a sensitive and reliable method for mRNA quantification and rapid analysis of gene expression from a large number of starting templates. It is based on the statistical significance of the beginning of exponential phase in real-time PCR kinetics, reflecting quantitative cycle of the initial target quantity and the efficiency of the PCR reaction (the fold increase of product per cycle). Results We used the large clinical biomarker dataset and 94-replicates-4-dilutions set which was published previously as research tools, then proposed a new qPCR curve analysis method——CqMAN, to determine the position of quantitative cycle as well as the efficiency of the PCR reaction and applied in the calculations. To verify algorithm performance, 20 genes from biomarker and partial data with concentration gradients from 94-replicates-4-dilutions set of MYCN gene were used to compare our method with various publicly available methods and established a suitable evaluation index system. Conclusions The results show that CqMAN method is comparable to other methods and can be a feasible method which applied to our self-developed qPCR data processing and analysis software, providing a simple tool for qPCR analysis.

49 Predicting Breeding Values of Purebred Pigs for Crossbred Performance Using Crossbred Phenotypes and Genotypes

Phenotyping a large number of crossbred progeny for the evaluation of purebred animals can be expensive. As genotyping with low-density panels is becoming cheaper, we aimed to evaluate the tradeoff between having different percentages of genotypes and phenotypes for crossbred progeny of candidate boars. We used the linear regression (LR) method to investigate changes in accuracy, bias, and inflation of breeding values for crossbred traits in purebred boars. A total of 304,582 purebred and 147,474 crossbred animals were phenotyped for average daily gain (ADG) and backfat thickness (BF), out of which 46,691 purebred and 13,117 crossbred animals were genotyped. Genomic information consisted of imputed genotypes for 40,247 SNP markers after quality control. A four-trait animal model under single-step GBLUP was used that included phenotypes recorded in purebred and crossbred animals as correlated traits. The LR statistics were calculated based on breeding values of young purebred sires from complete and partial data. The first complete data included genotypes for purebreds and phenotypes for purebreds and crossbreds, whereas the second included also genotypes for crossbreds. The partial data included phenotypes on 50% or none of the progeny of validation sires, with or without genotypes for crossbred animals. When 50% of the progeny has phenotypes, adding genotypes for crossbred progeny marginally increased accuracy of ADG (0.77 vs 0.78) for 47 boars with more than 150 progeny with phenotypes. No increase was observed for BF. A small increase in bias and inflation by adding crossbred genotypes was observed for ADG but not for BF. When no phenotypes were available for crossbred progeny, accuracy for both traits was lower but improved with crossbred genotypes for ADG (0.61 vs 0.64) for boars with more than 150 progeny. The tradeoff between phenotypes and genotypes should be further investigated in larger datasets with more validation boars.

Literacy skills seem to fuel literacy enjoyment, rather than vice versa

Children who like to read and write tend to be better at it. This association is typically interpreted as enjoyment impacting engagement in literacy activities, which boosts literacy skills. We fitted direction-of-causation models to partial data of 3,690 Finnish twins aged 12. Literacy skills were rated by the twins’ teachers and literacy enjoyment by the twins themselves. A bivariate twin model showed substantial genetic influences on literacy skills (70%) and literacy enjoyment (35%). In both skills and enjoyment, shared-environmental influences explained about 20% in each. Direction-of-causation modelling showed that skills impacted enjoyment. The influence in the other direction was zero. The genetic influences on skills influenced enjoyment, via the skills--> enjoyment path. This indicates active gene-environment correlation: children with an aptitude for good literacy skills are more likely to seek out literacy activities. To a lesser extent, it was also the shared-environmental influences on children’s skills that propagated to influence children’s literacy enjoyment. Environmental influences that foster children’s literacy skills (e.g., families and schools), also foster children’s love for reading and writing. These findings underline the importance of nurturing children’s literacy skills.

Pseudo-Likelihood Estimation for Parameters of Stochastic Time-Fractional Diffusion Equations

Although stochastic fractional partial differential equations have received increasing attention in the last decade, the parameter estimation of these equations has been seldom reported in literature. In this paper, we propose a pseudo-likelihood approach to estimating the parameters of stochastic time-fractional diffusion equations, whose forward solver has been investigated very recently by Gunzburger, Li, and Wang (2019). Our approach can accurately recover the fractional order, diffusion coefficient, as well as noise magnitude given the discrete observation data corresponding to only one realization of driving noise. When only partial data is available, our approach can also attain acceptable results for intermediate sparsity of observation.

Examination of New York City Transit’s Bus and Subway Ridership Trends During the COVID-19 Pandemic

The New York City metropolitan area was hard hit by COVID-19, and the pandemic brought with it unprecedented challenges for New York City Transit. This paper addresses the techniques used to estimate dramatically changing ridership, at a time when previously dependable sources suddenly became unavailable (e.g., local bus payment data, manual field checks). The paper describes alterations to ridership models, as well as the expanding use of automated passenger counters, including validation of new technology and scaling to account for partial data availability. The paper then examines the trends in subway and bus ridership. Peak periods shifted by both time of day and relative intensity compared with the rest of the day, but not in the same way on weekdays and weekends. On average, trip distances became longer for subway and local bus routes, but overall average bus trip distances decreased owing to a drop in express bus usage. Subway ridership changes were compared with neighborhood demographic statistics and numerous correlations were identified, including with employment, income, and race and ethnicity. Other factors, such as the presence of hospitals, were not found to be significant.

XStore : Fast RDMA-Based Ordered Key-Value Store Using Remote Learned Cache

RDMA ( Remote Direct Memory Access ) has gained considerable interests in network-attached in-memory key-value stores. However, traversing the remote tree-based index in ordered key-value stores with RDMA becomes a critical obstacle, causing an order-of-magnitude slowdown and limited scalability due to multiple round trips. Using index cache with conventional wisdom—caching partial data and traversing them locally—usually leads to limited effect because of unavoidable capacity misses, massive random accesses, and costly cache invalidations. We argue that the machine learning (ML) model is a perfect cache structure for the tree-based index, termed learned cache . Based on it, we design and implement XStore , an RDMA-based ordered key-value store with a new hybrid architecture that retains a tree-based index at the server to perform dynamic workloads (e.g., inserts) and leverages a learned cache at the client to perform static workloads (e.g., gets and scans). The key idea is to decouple ML model retraining from index updating by maintaining a layer of indirection from logical to actual positions of key-value pairs. It allows a stale learned cache to continue predicting a correct position for a lookup key. XStore ensures correctness using a validation mechanism with a fallback path and further uses speculative execution to minimize the cost of cache misses. Evaluations with YCSB benchmarks and production workloads show that a single XStore server can achieve over 80 million read-only requests per second. This number outperforms state-of-the-art RDMA-based ordered key-value stores (namely, DrTM-Tree, Cell, and eRPC+Masstree) by up to 5.9× (from 3.7×). For workloads with inserts, XStore still provides up to 3.5× (from 2.7×) throughput speedup, achieving 53M reqs/s. The learned cache can also reduce client-side memory usage and further provides an efficient memory-performance tradeoff, e.g., saving 99% memory at the cost of 20% peak throughput.