This paper addresses the trade-off between Accuracy and Transparency for deep learning applied to sports analytics. Neural nets achieve great predictive accuracy through deep learning, and are popular in sports analytics. But it is hard to interpret a neural net model and harder still to extract actionable insights from the knowledge implicit in it. Therefore, we built a simple and transparent model that mimics the output of the original deep learning model and represents the learned knowledge in an explicit interpretable way. Our mimic model is a linear model tree, which combines a collection of linear models with a regression-tree structure. The tree version of a neural network achieves high fidelity, explains itself, and produces insights for expert stakeholders such as athletes and coaches. We propose and compare several scalable model tree learning heuristics to address the computational challenge from datasets with millions of data points.
Abstract In response to the shortcomings of magnesium oxychloride cement (MOC), which has excellent with outstanding mechanical properties and favorable bioactivity but poor water resistance, strontium hydrogen phosphate (SrHPO 4 ) was used as a water resistance modifier for MOC in the paper, and the effects of SrHPO 4 on the strength, water resistance, in vitro degradation and bioactivity of MOC were investigated; the results showed that SrHPO 4 could enhance the water resistance of MOC, in which the initial strength of MOC containing 4.0 wt% SrHPO 4 was 92.3 ± 2.5 MPa, and the strength was still 8.2 ± 0.9 MPa after 84 d of immersion in SBF solution; the degradation experiments of the samples in SBF solution showed that the degradation of SrHPO 4 -MOC was controlled, and the low alkaline environment created by the degradation promoted the deposition of hydroxyapatite on the cement surface, it indicated that SrHPO 4 -MOC material had good degradation properties and bioactivity; cell experiments showed that compared with MOC, SrHPO 4 -MOC was noncytotoxic and could promote cell proliferation, which was expected to be a new material for bone repair.
Abstract Purpose: This paper explores the effect of active break on the distinguishability of autonomic nervous patterns of learning states through machine learning. Method: First, we collected electrocardiogram (ECG) data of 77 subjects before and after active break, and accurately located the R-wave peaks from the ECG signal to calculate the RR interval series. Second, the RR interval samples were segmented according to the inclusion criteria of certain learning states. The initial 39 ECG features were empirically calculated, and the optimal feature combination for learning states recognition was selected through sequential backward selection and leave-one-subject-out cross test. Finally, we established binary-classification models of pairs of learning states and compared their performance of learning states recognition before and after active break. Discussion: Active break can promote or inhibit the students’ academic performance. Besides statistical analysis shows the stability of the cognitive ability. What’s more active break increases the physiological response to fatigue Conclusion: (1) The autonomic nervous patterns of knowledge input-processing and retrieval-processing and those of cognitive load matching and mismatching in knowledge retrieval processing became less distinguishable with the effect of active break. (2) The autonomic nervous patterns of mental fatigue and no fatigue states became more distinguishable with the effect of active break. (3) Stronger parasympathetic nervous activities make students achieve better academic performance during using new knowledge to solve problem stage.
A novel nano-hydroxyapatite/poly(desaminotyrosyl-tyrosin-hexyl ester) carbonate composites bone cement was prepared using nano-hydroxyapatite and poly(desaminotyrosyl-tyrosin-hexyl ester) carbonate. The experimental results indicate that the compressive strength of the nano-hydroxyapatite/poly(desaminotyrosyl-tyrosinhexyl ester) carbonate composite bone cement which contained bone can reach 18.2 MPa, which can satisfy the repair demand of cancellous bone parts. Nano-hydroxyapatite and poly(desaminotyrosyl-tyrosinhexyl ester) carbonate are combined by hydrogen bonding. Calcium citrate is formed by reaction of Ca 2+ in nano-hydroxyapatite with the carboxyl group in citric acid. The nano-hydroxyapatite/poly(desaminotyrosyltyrosin-hexyl ester) carbonate composite bone cement was soaked in simulated body fluid to form hydroxyapatite. The results of the cell culture experiment and animal implantation indicated that nano-hydroxyapatite/poly(desaminotyrosyl-tyrosin-hexyl ester) carbonate composite bone cement had good biocompatibility and osteogenesis capability. This study is useful for the development of nanohydroxyapatite/ poly(desaminotyrosyl-tyrosin-hexyl ester) carbonate composite materials in the field of bone repair.
Squid are commercial marine species that have high nutritional value. This study aimed to compare the influences of vacuum frying and atmospheric frying on the physicochemical properties and protein oxidation of three main parts (ring, tentacle, and fin) of the squid Loligo chinensis. The results showed that the vacuum-fried (VF) group had lower moisture and total fat contents and looser microstructures than the atmospheric-fried (AF) group. The amino acid contents and molecular weight revealed that the proteins were well preserved during vacuum frying. Carbonyl content in the VF ring, tentacle, and fin samples increased nearly 2.53-, 1.54-, and 2.56-fold, respectively, compared to that in the corresponding fresh group, but these increases were lower than those of the corresponding AF group. In addition, the secondary structures of proteins revealed a slight decrease in the α-helix and β-turn contents and a significant increase in the β-sheet content during vacuum frying. Therefore, vacuum frying can be used as an efficient processing method to conserve the high nutritive quality of the product. PRACTICAL APPLICATION: As a developing alternative technology to prepare healthier fried products, vacuum frying has been the focus of recent researches. Vacuum frying produced squid products that had lower TBARS values, carbonyl contents, and Schiff base substances compared to atmospheric frying. And the protein secondary structures of the vacuum-fried group retained better. The study proved that vacuum frying could be an effective method with the advantages of high protein stability and product quality.
The current study aimed to determine the dynamic changes of plasma lutein and its kinetics profiles after a single oral administration of lutein, as well as the effect of omega‐3 fatty acids on lutein absorption in healthy Chinese people. Forty volunteers were randomly assigned to four dosage groups: 10mg lutein (Group L10), 20mg lutein (Group L20), 40mg lutein (Group L40), and 20mg lutein plus 560mg omega‐3 fatty acid (Group LF20). Fasting venous blood samples were collected at pre‐dosing and 4, 12, 24, 48, 72, 120, 168, 240, 360 h of post‐dosing in four groups. Plasma lutein concentrations at different time points were measured by high performance liquid chromatography (HPLC) for kinetics analysis. We found that plasma lutein concentrations in Group L40 were significantly higher than Group L10 at 12h, 24h, 48h ( p <0.01) and 72h post‐dosing ( p <0.05). The AUC 0‐360h , AUC 0‐∞ and C max were significantly higher in Group L40 than Group L10 and Group L20 ( p <0.01) and the results showed a dose‐dependent linear trend. The T max of plasma lutein was at 12h. There were no adverse events for a single supplementation with 40mg of lutein. In summary, our results showed that the absorption of lutein was positively correlated with dosage, and a single co‐administration of omega‐3 fatty acids did not affect the absorption of lutein. Support or Funding Information Supported by the National Key Technology Support Program (No. 2012BAD33B01‐2)
