Using Dynamic Multi-Task Non-Negative Matrix Factorization to Detect the Evolution of User Preferences in Collaborative Filtering

Predicting what items will be selected by a target user in the future is an important function for recommendation systems. Matrix factorization techniques have been shown to achieve good performance on temporal rating-type data, but little is known about temporal item selection data. In this paper, we developed a unified model that combines Multi-task Non-negative Matrix Factorization and Linear Dynamical Systems to capture the evolution of user preferences. Specifically, user and item features are projected into latent factor space by factoring co-occurrence matrices into a common basis item-factor matrix and multiple factor-user matrices. Moreover, we represented both within and between relationships of multiple factor-user matrices using a state transition matrix to capture the changes in user preferences over time. The experiments show that our proposed algorithm outperforms the other algorithms on two real datasets, which were extracted from Netflix movies and Last.fm music. Furthermore, our model provides a novel dynamic topic model for tracking the evolution of the behavior of a user over time.     

Emergence of scale-free leadership structure in social recommender systems

The study of the organization of social networks is important for the understanding of opinion formation, rumor spreading, and the emergence of trends and fashion. This paper reports empirical analysis of networks extracted from four leading sites with social functionality (Delicious, Flickr, Twitter and YouTube) and shows that they all display a scale-free leadership structure. To reproduce this feature, we propose an adaptive network model driven by social recommending. Artificial agent-based simulations of this model highlight a "good get richer" mechanism where users with broad interests and good judgments are likely to become popular leaders for the others. Simulations also indicate that the studied social recommendation mechanism can gradually improve the user experience by adapting to tastes of its users. Finally we outline implications for real online resource-sharing systems.     

Information Filtering via Heterogeneous Diffusion in Online Bipartite Networks

The rapid expansion of Internet brings us overwhelming online information, which is impossible for an individual to go through all of it. Therefore, recommender systems were created to help people dig through this abundance of information. In networks composed by users and objects, recommender algorithms based on diffusion have been proven to be one of the best performing methods. Previous works considered the diffusion process from user to object, and from object to user to be equivalent. We show in this work that it is not the case and we improve the quality of the recommendation by taking into account the asymmetrical nature of this process. We apply this idea to modify the state-of-the-art recommendation methods. The simulation results show that the new methods can outperform these existing methods in both recommendation accuracy and diversity. Finally, this modification is checked to be able to improve the recommendation in a realistic case.     

Trunk acceleration for neuroprosthetic control of standing: a pilot study

This pilot study investigated the potential of using trunk acceleration feedback control of center of pressure (COP) against postural disturbances with a standing neuroprosthesis following paralysis. Artificial neural networks (ANNs) were trained to use three-dimensional trunk acceleration as input to predict changes in COP for able-bodied subjects undergoing perturbations during bipedal stance. Correlation coefficients between ANN predictions and actual COP ranged from 0.67 to 0.77. An ANN trained across all subject-normalized data was used to drive feedback control of ankle muscle excitation levels for a computer model representing a standing neuroprosthesis user. Feedback control reduced average upper-body loading during perturbation onset and recovery by 42% and peak loading by 29% compared with optimal, constant excitation.     

An algorithm to classify amino acid sequences into protein groups of Bothrops jararacussu venomous gland

An algorithm for automatic clustering of database protein sequences from Bothrops jararacussu venomous gland, according to sequence similarities of their domains, is described. The program was written in C and Perl languages. This algorithm compares a domain with each ORF protein sequence in the database. Each nucleotide FASTA sequence generates six ORFs. As a result, the user has a list containing all sequences found in a specific domain and a display of the sequence, domain and number of hits. The algorithm lists only the sequences that present a minimum similarity of 30 hits and the best alignment. This limit was considered appropriate. The algorithm is available in the Internet (www.compbionet.org.br/cgi-domains/homesnake) and it can quickly and accurately organizes large database into classes.     

Item Anomaly Detection Based on Dynamic Partition for Time Series in Recommender Systems

In recent years, recommender systems have become an effective method to process information overload. However, recommendation technology still suffers from many problems. One of the problems is shilling attacks-attackers inject spam user profiles to disturb the list of recommendation items. There are two characteristics of all types of shilling attacks: 1) Item abnormality: The rating of target items is always maximum or minimum; and 2) Attack promptness: It takes only a very short period time to inject attack profiles. Some papers have proposed item anomaly detection methods based on these two characteristics, but their detection rate, false alarm rate, and universality need to be further improved. To solve these problems, this paper proposes an item anomaly detection method based on dynamic partitioning for time series. This method first dynamically partitions item-rating time series based on important points. Then, we use chi square distribution (χ²) to detect abnormal intervals. The experimental results on MovieLens 100K and 1M indicate that this approach has a high detection rate and a low false alarm rate and is stable toward different attack models and filler sizes.     

Recommended Best Practices for Process Monitoring Instrumentation in Pharmaceutical Freeze Drying—2017

Recommended best practices in monitoring of product status during pharmaceutical freeze drying are presented, focusing on methods that apply to both laboratory and production scale. With respect to product temperature measurement, sources of uncertainty associated with any type of measurement probe are discussed, as well as important differences between the two most common types of temperature-measuring instruments—thermocouples and resistance temperature detectors (RTD). Two types of pressure transducers are discussed—thermal conductivity-type gauges and capacitance manometers, with the Pirani gauge being the thermal conductivity-type gauge of choice. It is recommended that both types of pressure gauge be used on both the product chamber and the condenser for freeze dryers with an external condenser, and the reasoning for this recommendation is discussed. Developing technology for process monitoring worthy of further investigation is also briefly reviewed, including wireless product temperature monitoring, tunable diode laser absorption spectroscopy at manufacturing scale, heat flux measurement, and mass spectrometry as process monitoring tools.     

Characteristics and needs of different user groups in the urban forest of Stuttgart

In the urban forest of Stuttgart the forest roads are used by a growing number of people for recreational and leisure activities. To manage and channel the different user groups within an optimised forest road network needs a lot of information: Number and composition of visitors, demands of different user groups concerning standard of roads and trails they use for their activities, conflicts that may rise between different users using the same road at the same time, dedication of roads to special activities (e.g. fitness trail, hiking trail).

A relatively new method of video monitoring was used to collect long-term information about number and composition of visitors, personal interviews with “experts” were made to obtain information about the specific demands of the different user groups, and analysis of road network was done with the help of GIS. The results show that these tools complement one another quite well and the combination of obtained data may help to channel visitor flows and to minimise conflicts between different user groups.     

MatchMiner: a tool for batch navigation among gene and gene product identifiers

MatchMiner is a freely available program package for batch navigation among gene and gene product identifier types commonly encountered in microarray studies and other forms of 'omic' research. The user inputs a list of gene identifiers and then uses the Merge function to find the overlap with a second list of identifiers of either the same or a different type or uses the LookUp function to find corresponding identifiers.     

Managing visitor access to seabird colonies: a spatial simulation and empirical observations

Managers of wildlife reserves have a range of tools available to them when considering the best way to provide visitor access while avoiding as many of the negative effects of human disturbance as possible. However, managers lack guidelines as to whether conservation interests are best met by spreading visitors thinly throughout a reserve or by aggregating them in a small area. Here I describe how relationships between disturbance impact and disturbance pressure (the dose–response curve) can be used to address this issue. I generate a spatial simulation of two different models of visitor distribution (one more aggregated than the other) and explicitly model disturbance impact for a variety of dose–response curves. I show that the optimal visitor distribution is likely to depend on the sensitivity of the species and the overall visitor pressure. Importantly, I find that in certain circumstances optimal management can shift from one management option to the other if visitor numbers cross a certain threshold. I use published relationships predicting nesting success of Common Guillemots Uria aalge and Black-legged Kittiwakes Rissa tridactyla to assess optimal management at three nature reserves in Scotland. Optimal management for Guillemots depends on the number of people and the distance between the people and the birds. At sites with high disturbance pressures, management should aim to aggregate visitors in as small an area as possible, whereas at sites with lower disturbance pressure, an even distribution of visitors is favoured. Kittiwake models were not generally accurate, and consequently only site-specific guidelines could be generated, where an even distribution was favoured.     

Phosphate-haemoglobin interaction. The primary structure of the haemoglobin of the African elephant (Loxodonta africana, Proboscidea): asparagine in position 2 of the beta-chain

The primary structure of the haemoglobin of the African Elephant (Loxodonta africana) is reported. The sequence was determined by means of a sequenator. The haemoglobin differs in 26 amino acids in the alpha-chains and in 27 in the beta-chains from that of adult human haemoglobin. The haemoglobin of the African Elephant, like that of the Indian Elephant and Ilama, has only 5 binding sites for polyphosphate. This finding explains the low p(O2)50 value in whole blood as a result of the lower 2,3-bisphosphoglycerate-haemoglobin interaction. This is discussed in relation to aspects of respiratory physiology; some points are also of interest with regard to the Second Punic War and Hannibal's crossing of the Alps.     

Get It While It’s Hot: A Peak-First Bias in Self-Generated Choice Order in Rhesus Macaques

Animals typically must make a number of successive choices to achieve a goal: e.g., eating multiple food items before becoming satiated. However, it is unclear whether choosing the best first or saving the best for last represents the best choice strategy to maximize overall reward. Specifically, since outcomes can be evaluated prospectively (with future rewards discounted and more immediate rewards preferred) or retrospectively (with prior rewards discounted and more recent rewards preferred), the conditions under which each are used remains unclear. On the one hand, humans and non-human animals clearly discount future reward, preferring immediate rewards to delayed ones, suggesting prospective evaluation; on the other hand, it has also been shown that a sequence that ends well, i.e., with the best event or item last, is often preferred, suggesting retrospective evaluation. Here we hypothesized that when individuals are allowed to build the sequence themselves they are more likely to evaluate each item individually and therefore build a sequence using prospective evaluation. We examined the relationship between self-generated choice order and preference in rhesus monkeys in two experiments in which the distinctiveness of options were relatively high and low, respectively. We observed a positive linear relationship between choice order and preference among highly distinct options, indicating that the rhesus monkeys chose their preferred food first: i.e., a peak-first order preference. Overall, choice order depended on the degree of relative preference among alternatives and a peak-first bias, providing evidence for prospective evaluation when choice order is self-generated.     

Analysis of distance-based protein structure prediction by deep learning in CASP13

This paper reports the CASP13 results of distance-based contact prediction, threading, and folding methods implemented in three RaptorX servers, which are built upon the powerful deep convolutional residual neural network (ResNet) method initiated by us for contact prediction in CASP12. On the 32 CASP13 FM (free-modeling) targets with a median multiple sequence alignment (MSA) depth of 36, RaptorX yielded the best contact prediction among 46 groups and almost the best 3D structure modeling among all server groups without time-consuming conformation sampling. In particular, RaptorX achieved top L/5, L/2, and L long-range contact precision of 70%, 58%, and 45%, respectively, and predicted correct folds (TMscore > 0.5) for 18 of 32 targets. Further, RaptorX predicted correct folds for all FM targets with >300 residues (T0950-D1, T0969-D1, and T1000-D2) and generated the best 3D models for T0950-D1 and T0969-D1 among all groups. This CASP13 test confirms our previous findings: (a) predicted distance is more useful than contacts for both template-based and free modeling; and (b) structure modeling may be improved by integrating template and coevolutionary information via deep learning. This paper will discuss progress we have made since CASP12, the strength and weakness of our methods, and why deep learning performed much better in CASP13.     

PISCES: a protein sequence culling server

PISCES is a public server for culling sets of protein sequences from the Protein Data Bank (PDB) by sequence identity and structural quality criteria. PISCES can provide lists culled from the entire PDB or from lists of PDB entries or chains provided by the user. The sequence identities are obtained from PSI-BLAST alignments with position-specific substitution matrices derived from the non-redundant protein sequence database. PISCES therefore provides better lists than servers that use BLAST, which is unable to identify many relationships below 40% sequence identity and often overestimates sequence identity by aligning only well-conserved fragments. PDB sequences are updated weekly. PISCES can also cull non-PDB sequences provided by the user as a list of GenBank identifiers, a FASTA format file, or BLAST/PSI-BLAST output.     

Prediction of interresidue contacts with DeepMetaPSICOV in CASP13

In this article, we describe our efforts in contact prediction in the CASP13 experiment. We employed a new deep learning-based contact prediction tool, DeepMetaPSICOV (or DMP for short), together with new methods and data sources for alignment generation. DMP evolved from MetaPSICOV and DeepCov and combines the input feature sets used by these methods as input to a deep, fully convolutional residual neural network. We also improved our method for multiple sequence alignment generation and included metagenomic sequences in the search. We discuss successes and failures of our approach and identify areas where further improvements may be possible. DMP is freely available at: https://github.com/psipred/DeepMetaPSICOV .     

A large database DNA sequence handling program with generalized searching specifications.

The program described allows for the creation and manipulation of files of DNA sequence data up to very great lengths. The program uses its own paging system to load segments of the sequence into a small internal buffer so that the program does not have excessive memory requirements. The program offers a menu of functions to the user, and has been written to be forgiving of user errors. A code for the generalised specification of bases as a series of groups (i.e. A or T, Purine, etc.) has been devised and can be used in search specifications or in sequence files. Versions of the program have been developed to run with special efficiency under DIGITAL's RT11 operating system or to run under systems with a suitable implementation of FORTRAN VI.