Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

Recent advancements in deep neural networks for graph-structured data have led to state-of-the-art performance on recommender system benchmarks. However, making these methods practical and scalable to web-scale recommendation tasks with billions of items and hundreds of millions of users remains an unsolved challenge. Here we describe a large-scale deep recommendation engine that we developed and deployed at Pinterest. We develop a data-efficient Graph Convolutional Network (GCN) algorithm, which combines efficient random walks and graph convolutions to generate embeddings of nodes (i.e., items) that incorporate both graph structure as well as node feature information. Compared to prior GCN approaches, we develop a novel method based on highly efficient random walks to structure the convolutions and design a novel training strategy that relies on harder-and-harder training examples to improve robustness and convergence of the model. We also develop an efficient MapReduce model inference algorithm to generate embeddings using a trained model. Overall, we can train on and embed graphs that are four orders of magnitude larger than typical GCN implementations. We show how GCN embeddings can be used to make high-quality recommendations in various settings at Pinterest, which has a massive underlying graph with 3 billion nodes representing pins and boards, and 17 billion edges. According to offline metrics, user studies, as well as A/B tests, our approach generates higher-quality recommendations than comparable deep learning based systems. To our knowledge, this is by far the largest application of deep graph embeddings to date and paves the way for a new generation of web-scale recommender systems based on graph convolutional architectures.

Graph Convolutional Neural Networks for Web-Scale Recommender Systems

https://www.cell.com/trends/neurosciences/pdf/S0166-2236(09)00153-2.pdf

Molecular dissection of reactive astrogliosis and glial scar formation.

Matrix Factorization Techniques for Recommender Systems

Theranostics: Combining Imaging and Therapy

Objective To investigate music appreciation in hearing aid (HA) users with varying levels of hearing loss (HL). Study Design Cross-sectional, within-subjects design. Setting Tertiary medical center, community. Patients Adults (≥18 yr) bilateral HA users. Interventions HA usage. Main Outcome Measures Outcome variables included self-reported music enjoyment measures (pleasantness, musicality, naturalness) with and without HAs assessed with visual analogue scales (10 indicates highest level of enjoyment, 0 the least). Exposure variables include HL (better ear pure-tone average) and speech discrimination (word recognition scores [WRS]). Demographic information was collected. Results One hundred nine bilateral HA users completed the study. Mean (standard deviation) age was 66.6 years (16.8 yr); 52.3% were female patients. Mean (standard deviation) better ear pure-tone average was 51.1 dB (16.3 dB) HL. Increased severity of HL and worse WRS were associated with decreased music enjoyment (p < 0.05) across all measures without HAs, adjusting for sex, age, education, race, HA type, age of HL diagnosis, duration of HL, duration of HA use, musical preference, and musical experience. However, these associations were attenuated or no longer significant with HA usage. Moreover, among all subjects, HAs (vs. no HAs) provided increased music enjoyment in pleasantness (HA, 6.94, no HA, 5.74; p < 0.01), musicality (HA, 7.35; no HA, 6.13, p < 0.01), and naturalness (HA, 6.75; no HA, 6.02; p = 0.02). Conclusion HA users report increased music enjoyment with HAs compared with without HAs. Increased severity of HL and worse WRS were independently associated with decreased unaided music enjoyment. HA usage seems to mitigate this effect, particularly for those with worse HL.

Hearing Aids Enhance Music Enjoyment in Individuals With Hearing Loss

Facial Plastic Surgery & Aesthetic MedicineAhead of Print Differences in Patient Characteristics and Spending Among Individuals Undergoing Gender-Affirming Rhinoplasty in the United States from 2016 to 2019Rishub Karan Das, Rahul K. Sharma, Salam Al Kassis, Brian C. Drolet, Galen Perdikis, and Priyesh N. PatelRishub Karan Das*Address correspondence to: Rishub Karan Das, BA, Vanderbilt University School of Medicine, 1161 21st Avenue South, Nashville, TN 37212, USA. E-mail Address: rishub.das@vanderbilt.eduhttps://orcid.org/0000-0002-7368-7769Vanderbilt University School of Medicine, Nashville, Tennessee, USA.Search for more papers by this author, Rahul K. Sharmahttps://orcid.org/0000-0003-3120-6194Department of Otolaryngology, Vanderbilt University Medical Center, Nashville, Tennessee, USA.Search for more papers by this author, Salam Al KassisDepartment of Plastic Surgery, Vanderbilt University Medical Center, Nashville, Tennessee, USA.Search for more papers by this author, Brian C. DroletDepartment of Plastic Surgery, Vanderbilt University Medical Center, Nashville, Tennessee, USA.Search for more papers by this author, Galen PerdikisDepartment of Plastic Surgery, Vanderbilt University Medical Center, Nashville, Tennessee, USA.Search for more papers by this author, and Priyesh N. Patelhttps://orcid.org/0000-0002-5850-2172Department of Otolaryngology, Vanderbilt University Medical Center, Nashville, Tennessee, USA.Search for more papers by this authorPublished Online:10 Jan 2023https://doi.org/10.1089/fpsam.2022.0315AboutSectionsView articleView Full TextSupplemental MaterialPDF/EPUBView Supplemental Data Permissions & CitationsDownload CitationsTrack CitationsAdd to favorites Back To Publication ShareShare onFacebookTwitterLinked InRedditEmail View articleFiguresReferencesRelatedDetails Volume 0Issue 0 InformationCopyright 2023, American Academy of Facial Plastic and Reconstructive Surgery, Inc.To cite this article:Rishub Karan Das, Rahul K. Sharma, Salam Al Kassis, Brian C. Drolet, Galen Perdikis, and Priyesh N. Patel.Differences in Patient Characteristics and Spending Among Individuals Undergoing Gender-Affirming Rhinoplasty in the United States from 2016 to 2019.Facial Plastic Surgery & Aesthetic Medicine.ahead of printhttp://doi.org/10.1089/fpsam.2022.0315Online Ahead of Print:January 10, 2023PDF download 

Differences in Patient Characteristics and Spending Among Individuals Undergoing Gender-Affirming Rhinoplasty in the United States from 2016 to 2019.

This study explores the synthesis, properties, and applications of chitosan-encapsulated silver sulphide (Ag2S) quantum dots (QDs) for biological applications. The investigation focuses on the fluctuations in the physico-chemical characteristics of chitosan Ag2S QDs, which can be carefully studied due to their environmental activity. X-ray diffraction (XRD) measurements reveal that chitosan-coated Ag2S QDs exhibit higher-intensity peaks. The XRD analysis reports a range of crystallite sizes, with a minimum size of 8 nm and a maximum size of 12 nm. Fourier-transform infrared (FTIR) spectroscopy confirms the presence of chitosan through the detection of functional group peaks. High-resolution transmission electron microscopy (HRTEM) studies indicate that the size of the artificial quantum dots is 6 nm. Energy-dispersive X-ray spectroscopy (EDX) verifies the composition of chitosan-encapsulated Ag2S QDs. Moreover, the chitosan Ag2S quantum dots demonstrate exceptional photocatalytic activity, as evidenced by the degradation of 92% of methylene blue dye within one hour. This research provides valuable insights into the synthesis, properties, and potential applications of chitosan-encapsulated Ag2S quantum dots in diverse fields.

Enhanced Photocatalytic Applications of Chitosan Encapsulated Silver Sulphide Quantum Dots

Background: Cancer chemotherapy is known to be accompanied by adverse drug reactions (ADRs). There is a paucity of documented studies concerning the factors predisposing to these ADRs in the English medical literature. Objective: To evaluate the role played by socio-demographic factors in predisposition to ADRs in patients on cancer chemotherapy. Material and Methods: In this prospective and observational study, outcome of relationship between socio-demographic factors and ADRs occurring in patients receiving cancer chemotherapy in a regional oncology center in Jammu and Kashmir, India over a period of one year was studied. Results: Overall, 90% of the 102 cancer patients expressed ADRs to chemotherapy. As many as 268 ADRs were reported in the males (48.53 ± 15.46 years) was slightly higher as compared to that of females (45.53 ± 14.89 years). Aged patients had higher frequency of ADRs than young ones. Females were observed to express ADRs more often than males. Illiterate patients expressed more ADRs compared to educated patients. Conclusion: Knowledge of socio-demographic factors may assist clinicians to address these factors in prediction of predisposition to ADRs in cancer chemotherapy.

Journal of Cancer Therapy and Research Cancer Chemotherapy-Induced Adverse Drug Reactions: A Prospective Study of Role of Socio-Demographic Factors

Pannexin channel isoforms (Panx1‐3) are thought to release nucleotides into the extracellular milieu and have been shown to effect vascular hemodynamics. For this reason, we examined their mRNA and protein expression in hypertensive humans and genetically‐inbred hypertensive mice. In both mouse and humans, we found a significant reduction in Panx3 expression in resistance artery endothelium. Thus, we hypothesized Panx3 may be a regulator of vascular function. In en face endothelial preparations from 3rd order mesenteric arteries, we localized Panx3 to the Golgi Apparatus as opposed to Panx1 which localized to the plasma membrane. Next, we generated an inducible, endothelial cell Panx3 knockout mouse (Panx3ECKO). Radiotelemetry revealed a renin‐independent spontaneous hypertension, with unremarkable immune infiltration in the kidney. There was no change in cytoplasmic or released ATP. To understand how Panx3 may regulate blood pressure, we examined whether Panx3 interacted with B Cell Lymphoma 6 (BCL6), a potential binding partner. En face proximity ligation assays demonstrated an interaction between Panx3 and BCL6 in the Golgi. Panx3ECKO mice exhibited significantly decreased BCL6 protein, hinting that Panx3 may stabilize BCL6 by binding at the BCL6 ubiquitin sites. In silico “threading” of the Panx3 sequence onto the cryo‐EM structure of Panx1 confirmed this site of interaction. BCL6 is a NFκB repressor, thus its degradation in Panx3ECKO mice caused an increase in NFκB activity with IκBα and p100 significantly upregulated. A novel mimetic peptide designed to block Panx3‐BCL6 interactions was administered into C57Bl/6J mice, which recapitulated these results. In addition, Panx3ECKO mice had increased endothelial NOX4 (but not NOX1 or NOX2), likely due to increased NFκB activity—this correlated with a significant increase in plasma H2O2, nitrotyrosine (NT3), and 4‐hydroxynonenal (4‐HNE). In line with this observation, 3rd order mesenteric arteries from Panx3ECKO mice constricted (not dilated) to acetylcholine, which was rescued with the H2O2‐scavenger catalase (1000U/mL), suggesting that vascular oxidative stress drives hypertension in Panx3ECKO mice. Interestingly, Panx3ECKO mice also exhibit a significant increase in IL‐4 receptors on endothelium, and increased IL‐4 cytokines in bone marrow lysates. Because IL‐4 can drive BCL6 expression in other cell types, we suggest a possible homeostatic immune‐endothelial signaling axis. These data elucidate a novel Golgi‐localized oxidative signaling pathway in endothelium with a potential immune‐derived negative feedback loop.

Rahul Sharma

Papers

Hearing Aids Enhance Music Enjoyment in Individuals With Hearing Loss

Differences in Patient Characteristics and Spending Among Individuals Undergoing Gender-Affirming Rhinoplasty in the United States from 2016 to 2019.

Enhanced Photocatalytic Applications of Chitosan Encapsulated Silver Sulphide Quantum Dots

Journal of Cancer Therapy and Research Cancer Chemotherapy-Induced Adverse Drug Reactions: A Prospective Study of Role of Socio-Demographic Factors

Endothelial Pannexin 3 – B Cell Lymphoma‐6 Interactions Protect Against Oxidative Stress