Typically, learned robot controllers are trained via relatively unsystematic regimens and evaluated with coarse-grained outcome measures such as average cumulative reward. The typical approach is useful to compare learning algorithms but provides limited insight into the effects of different training regimens and little understanding about the richness and complexity of learned behaviors. Likewise, human infants and other animals are "trained" via unsystematic regimens, but in contrast, developmental psychologists evaluate their performance in highly-controlled experiments with fine-grained measures such as success, speed of walking, and prospective adjustments. However, the study of learned behavior in human infants is limited by the practical constraints of training and testing babies. Here, we present a case study that applies methods from developmental psychology to study the learned behavior of the simulated bipedal robot Cassie. Following research on infant walking, we systematically designed reinforcement learning training regimens and tested the resulting controllers in simulated environments analogous to those used for babies--but without the practical constraints. Results reveal new insights into the behavioral impact of different training regimens and the development of Cassie's learned behaviors relative to infants who are learning to walk. This interdisciplinary baby-robot approach provides inspiration for future research designed to systematically test effects of training on the development of complex learned robot behaviors.

Drug resistance presents a major challenge in cancer therapy. Single cell profiling offers insights into cellular heterogeneity, yet the application of large-scale foundation models for predicting drug response in single cell data remains underexplored. To address this, we developed scDrugMap, an integrated framework featuring both a Python command-line interface and a web server for drug response prediction. scDrugMap evaluates a wide range of foundation models, including eight single-cell models and two large language models, using a curated dataset of over 326,000 cells in the primary collection and 18,800 cells in the validation set, spanning 36 datasets and diverse tissue and cancer types. We benchmarked model performance under pooled-data and cross-data evaluation settings, employing both layer freezing and Low-Rank Adaptation (LoRA) fine-tuning strategies. In the pooled-data scenario, scFoundation achieved the best performance, with mean F1 scores of 0.971 (layer freezing) and 0.947 (fine-tuning), outperforming the lowest-performing model by over 50%. In the cross-data setting, UCE excelled post fine-tuning (mean F1: 0.774), while scGPT led in zero-shot learning (mean F1: 0.858). Overall, scDrugMap provides the first large-scale benchmark of foundation models for drug response prediction in single-cell data and serves as a user-friendly, flexible platform for advancing drug discovery and translational research.

Objective: Create precise, structured, data-backed guidelines for type 2 diabetes treatment progression, suitable for clinical adoption. Research Design and Methods: Our training cohort was composed of patient (with type 2 diabetes) visits from Boston Medical Center (BMC) from 1998 to 2014. We divide visits into 4 groups based on the patient's treatment regimen before the visit, and further divide them into subgroups based on the recommended treatment during the visit. Since each subgroup has observational data, which has confounding bias (sicker patients are prescribed more aggressive treatments), we used machine learning and optimization to remove some datapoints so that the remaining data resembles a randomized trial. On each subgroup, we train AI-backed tree-based models to prescribe treatment changes. Once we train these tree models, we manually combine the models for every group to create an end-to-end prescription pipeline for all patients in that group. In this process, we prioritize stepping up to a more aggressive treatment before considering less aggressive options. We tested this pipeline on unseen data from BMC, and an external dataset from Hartford healthcare (type 2 diabetes patient visits from January 2020 to May 2024). Results: The median HbA1c reduction achieved by our pipelines is 0.26% more than what the doctors achieved on the unseen BMC patients. For the Hartford cohort, our pipelines were better by 0.13%. Conclusions: This precise, interpretable, and efficient AI-backed approach to treatment progression in type 2 diabetes is predicted to outperform the current practice and can be deployed to improve patient outcomes.

Lay summarisation aims to jointly summarise and simplify a given text, thus making its content more comprehensible to non-experts. Automatic approaches for lay summarisation can provide significant value in broadening access to scientific literature, enabling a greater degree of both interdisciplinary knowledge sharing and public understanding when it comes to research findings. However, current corpora for this task are limited in their size and scope, hindering the development of broadly applicable data-driven approaches. Aiming to rectify these issues, we present two novel lay summarisation datasets, PLOS (large-scale) and eLife (medium-scale), each of which contains biomedical journal articles alongside expert-written lay summaries. We provide a thorough characterisation of our lay summaries, highlighting differing levels of readability and abstractiveness between datasets that can be leveraged to support the needs of different applications. Finally, we benchmark our datasets using mainstream summarisation approaches and perform a manual evaluation with domain experts, demonstrating their utility and casting light on the key challenges of this task.

In every field--health, education, marketing, production, and sports--AI has managed to outperform itself. The sports industry is being rapidly transformed by artificial intelligence, much like every other significant sector. AI has touched every sports aspect at the quantitative, statistical, and analytical levels. The sports market has grown significantly. Game nights have become very popular thanks to the crazed devotees who have brought in a lot of money.

U.S. researchers have used machine learning to predict the effectiveness of multi-drug treatment combinations for tuberculosis (TB), which could help in the design of new therapy regimens. By examining study data from TB drug pairs in vitro, they were able to predict how three or four drugs could affect treatment in vivo and work out rules governing drug choices among these pairwise building blocks that would create effective multi-drug therapies. "Using the design rules we've established and tested, we can substitute one drug pair for another drug pair and know with a high degree of confidence that the drug pair should work in concert with the other drug pair to kill the TB bacteria in the rodent model," explained researcher Bree Aldridge, associate professor of molecular biology and microbiology at Tufts University School of Medicine in Boston, Massachusetts. "The selection process we developed is both more streamlined and more accurate in predicting success than prior processes, which necessarily considered fewer combinations." The research, published in the journal Cell Reports Medicine, follows an earlier study released this week showing that a deep learning program can be as effective as radiologists in identifying tuberculosis on chest X-rays.

Despite major advances in treatment and diagnosis over the past decades, cancer still ranks as a leading cause of mortality and a major impediment to extending life expectancy worldwide. Artificial intelligence's future involvement in healthcare, particularly in the detection and treatment of cancer, is anticipated to take a variety of forms, ranging from identifying a specific form of cancer to evaluating which therapy method may best treat that particular instance. AI promises to increase customization of cancer care and help individuals live with the illness with a greater quality of life and fewer side effects. In order to discover cancer in its most treatable stage, screenings are meant to monitor patients who do not show any symptoms proactively. U.S. Preventative Services Task Force recommends screening for breast, cervical, colorectal, and lung cancer.

We establish a broad methodological foundation for mixed-integer optimization with learned constraints. We propose an end-to-end pipeline for data-driven decision making in which constraints and objectives are directly learned from data using machine learning, and the trained models are embedded in an optimization formulation. We exploit the mixed-integer optimization-representability of many machine learning methods, including linear models, decision trees, ensembles, and multi-layer perceptrons. The consideration of multiple methods allows us to capture various underlying relationships between decisions, contextual variables, and outcomes. We also characterize a decision trust region using the convex hull of the observations, to ensure credible recommendations and avoid extrapolation. We efficiently incorporate this representation using column generation and clustering. In combination with domain-driven constraints and objective terms, the embedded models and trust region define a mixed-integer optimization problem for prescription generation. We implement this framework as a Python package (OptiCL) for practitioners. We demonstrate the method in both chemotherapy optimization and World Food Programme planning. The case studies illustrate the benefit of the framework in generating high-quality prescriptions, the value added by the trust region, the incorporation of multiple machine learning methods, and the inclusion of multiple learned constraints.

The problem of scheduling chemotherapy treatments in oncology clinics is a complex problem, given that the solution has to satisfy (as much as possible) several requirements such as the cyclic nature of chemotherapy treatment plans, maintaining a constant number of patients, and the availability of resources, e.g., treatment time, nurses, and drugs. At the same time, realizing a satisfying schedule is of upmost importance for obtaining the best health outcomes. In this paper we first consider a specific instance of the problem which is employed in the San Martino Hospital in Genova, Italy, and present a solution to the problem based on Answer Set Programming (ASP). Then, we enrich the problem and the related ASP encoding considering further features often employed in other hospitals, desirable also in S. Martino, and/or considered in related papers. Results of an experimental analysis, conducted on the real data provided by the San Martino Hospital, show that ASP is an effective solving methodology also for this important scheduling problem.

AI models capable of reliably guiding driverless cars typically require endless testing and fine-tuning, not to mention computational power out the wazoo. In an effort to bolster AI algorithm training effectiveness and efficiency, Google parent company Alphabet's Waymo is collaborating with DeepMind on techniques inspired by evolutionary biology, the two companies revealed in a blog post this morning. As Waymo explains, AI algorithms self-improve through trial and error. A model is presented with a task that it learns to perform by continually attempting it and adjusting based on the feedback it receives. Performance is heavily dependent on the training regimen -- known as a hyperparemeter schedule -- and finding the best regimen is commonly left to experienced researchers and engineers. They handpick AI models undergoing training, culling the weakest performers and freeing resources to train new algorithms from scratch.