Contrastive Learning (CL) has achieved impressive performance in self-supervised learning tasks, showing superior generalization ability. Inspired by the success, adopting CL into collaborative filtering (CF) is prevailing in semi-supervised topK recommendations. The basic idea is to routinely conduct heuristic-based data augmentation and apply contrastive losses (e.g., InfoNCE) on the augmented views. Yet, some CF-tailored challenges make this adoption suboptimal, such as the issue of out-of-distribution, the risk of false negatives, and the nature of top-K evaluation. They necessitate the CL-based CF scheme to focus more on mining hard negatives and distinguishing false negatives from the vast unlabeled user-item interactions, for informative contrast signals.

There are 45 Prime Day tech deals under 50 that are still live for the second day of Amazon's sale. Here, we've gathered up the smaller gadgets and supporting accessories, brands and devices we've tested and reviewed and know to be worth your money. This list includes useful gear like chargers, storage cards, cables, batteries and even earbuds -- plus a few Lego sets thrown in for good measure. Here are the best Prime Day Tech deals under 50. If you've got 25 and some change in an account somewhere, you can get something decent from Amazon's sale (particularly if you're a Prime member and don't have to pay for shipping).

October Prime Day deals are proving to be just as good as those we saw in July, at least for some of Amazon's own devices. The recently revamped Echo Spot is back on sale for a record low of 45, a discount that we last saw during the summertime sale event. The original Echo Spot came out in 2017, but it only lasted two years before it was discontinued. Amazon brought back a redesigned version of the Alexa speaker earlier this year. You can also opt for a bundle and get a TP-Link Tapo Smart Color Bulb with it for the same price.

Collaborative Metric Learning (CML) has recently emerged as a popular method in recommendation systems (RS), closing the gap between metric learning and Collaborative Filtering. Following the convention of RS, existing methods exploit unique user representation in their model design. This paper focuses on a challenging scenario where a user has multiple categories of interests. Under this setting, we argue that the unique user representation might induce preference bias, especially when the item category distribution is imbalanced. To address this issue, we propose a novel method called Diversity-Promoting Collaborative Metric Learning (DPCML), with the hope of considering the commonly ignored minority interest of the user.

Psychological research shows that enjoyment of many goods is subject to satiation, with short-term satisfaction declining after repeated exposures to the same item. Nevertheless, proposed algorithms for powering recommender systems seldom model these dynamics, instead proceeding as though user preferences were fixed in time. In this work, we introduce rebounding bandits, a multi-armed bandit setup, where satiation dynamics are modeled as time-invariant linear dynamical systems. Expected rewards for each arm decline monotonically with consecutive exposures and rebound towards the initial reward whenever that arm is not pulled. Unlike classical bandit algorithms, methods for tackling rebounding bandits must plan ahead and model-based methods rely on estimating the parameters of the satiation dynamics.

Zoom's Zoomtopia 2024 feature drop is, perhaps unsurprisingly, focused on actively integrating AI even more into its platform, especially for Zoom Workplace and Zoom Business Services. There's plenty to talk about, but we'll focus on Zoom AI Companion 2.0 for Zoom Workplace since this appears to be the biggest news the company is announcing today. AI Companion 2.0 promises to be a persistent presence in a Zoom Workplace window, and it will even remember previous conversations and come up with citations for statements users made. Since it's connected to the web, it can work with external apps like Gmail, Google Calendar and Microsoft Office. Other functions include summarizing documents, email chains and generating content drafts.

Large-scale industrial recommender systems are usually confronted with computational problems due to the enormous corpus size. To retrieve and recommend the most relevant items to users under response time limits, resorting to an efficient index structure is an effective and practical solution. The previous work Tree-based Deep Model (TDM) \cite{zhu2018learning} greatly improves recommendation accuracy using tree index. By indexing items in a tree hierarchy and training a user-node preference prediction model satisfying a max-heap like property in the tree, TDM provides logarithmic computational complexity w.r.t. the corpus size, enabling the use of arbitrary advanced models in candidate retrieval and recommendation.

Thanks to their scalability, two-stage recommenders are used by many of today's largest online platforms, including YouTube, LinkedIn, and Pinterest. These systems produce recommendations in two steps: (i) multiple nominators--tuned for low prediction latency--preselect a small subset of candidates from the whole item pool; (ii) a slower but more accurate ranker further narrows down the nominated items, and serves to the user. Despite their popularity, the literature on two-stage recommenders is relatively scarce, and the algorithms are often treated as mere sums of their parts. Such treatment presupposes that the two-stage performance is explained by the behavior of the individual components in isolation. This is not the case: using synthetic and real-world data, we demonstrate that interactions between the ranker and the nominators substantially affect the overall performance.

We consider running multiple instances of multi-armed bandit (MAB) problems in parallel. A main motivation for this study are online recommendation systems, in which each of N users is associated with a MAB problem and the goal is to exploit users' similarity in order to learn users' preferences to K items more efficiently. We consider the adversarial MAB setting, whereby an adversary is free to choose which user and which loss to present to the learner during the learning process. Users are in a social network and the learner is aided by a-priori knowledge of the strengths of the social links between all pairs of users. It is assumed that if the social link between two users is strong then they tend to share the same action. The regret is measured relative to an arbitrary function which maps users to actions.

Plenty of the tech we cover costs less than 50. And some gadgets hovering close enough to that price just need a decent discount to put them in range. During Amazon's Prime Day sale, plenty of tech deals can be had for under 50 and we've gathered up the best of what's out there. If you need to pick up microSD cards, power banks, digital streamers or even a smart speaker, now's the time. As always, these Prime Day picks are drawn from our own testing, coverage and reviews.