https://www2.econ.iastate.edu/tesfatsi/DeepLearningInNeuralNetworksOverview.JSchmidhuber2015.pdf

Deep learning in neural networks

https://www.cs.jhu.edu/~misha/ReadingSeminar/Papers/Bay08.pdf

Speeded-Up Robust Features (SURF)

We will review some of the major results in random graphs and some of the more challenging open problems. We will cover algorithmic and structural questions. We will touch on newer models, including those related to the WWW.

Random graphs

/pdf/convex-analysisnoer-san-nojin-zhan-nituite-5dl2rbmoyr.pdf

Convex Analysisの二,三の進展について

Models for the processes by which ideas and influence propagate through a social network have been studied in a number of domains, including the diffusion of medical and technological innovations, the sudden and widespread adoption of various strategies in game-theoretic settings, and the effects of "word of mouth" in the promotion of new products. Recently, motivated by the design of viral marketing strategies, Domingos and Richardson posed a fundamental algorithmic problem for such social network processes: if we can try to convince a subset of individuals to adopt a new product or innovation, and the goal is to trigger a large cascade of further adoptions, which set of individuals should we target?We consider this problem in several of the most widely studied models in social network analysis. The optimization problem of selecting the most influential nodes is NP-hard here, and we provide the first provable approximation guarantees for efficient algorithms. Using an analysis framework based on submodular functions, we show that a natural greedy strategy obtains a solution that is provably within 63% of optimal for several classes of models; our framework suggests a general approach for reasoning about the performance guarantees of algorithms for these types of influence problems in social networks.We also provide computational experiments on large collaboration networks, showing that in addition to their provable guarantees, our approximation algorithms significantly out-perform node-selection heuristics based on the well-studied notions of degree centrality and distance centrality from the field of social networks.

/pdf/maximizing-the-spread-of-influence-through-a-social-network-47y1ehuvwi.pdf

Maximizing the Spread of Influence through a Social Network

We study "adversarial scaling", a multi-armed bandit model where rewards have a stochastic and an adversarial component. Our model captures display advertising where the "click-through-rate" can be decomposed to a (fixed across time) arm-quality component and a non-stochastic user-relevance component (fixed across arms). Despite the relative stochasticity of our model, we demonstrate two settings where most bandit algorithms suffer. On the positive side, we show that two algorithms, one from the action elimination and one from the mirror descent family are adaptive enough to be robust to adversarial scaling. Our results shed light on the robustness of adaptive parameter selection in stochastic bandits, which may be of independent interest.

Bandits with adversarial scaling

In online advertising, advertisers purchase ad placements by participating in a long sequence of repeated auctions. One of the most important features advertising platforms often provide, and advertisers often use, is budget management, which allows advertisers to control their cumulative expenditures. Advertisers typically declare the maximum daily amount they are willing to pay, and the platform adjusts allocations and payments to guarantee that cumulative expenditures do not exceed budgets. There are multiple ways to achieve this goal, and each one, when applied to all budget-constrained advertisers simultaneously, drives the system toward a different equilibrium. While previous research focused on online stochastic optimization techniques or game-theoretic equilibria of such settings, our goal is to compare the "system equilibria" of a range of budget management strategies. In particular, we consider six different budget management strategies including probabilistic throttling, thresholding, bid shading, reserve pricing, and two versions of multiplicative boosting. We show these methods admit a system equilibrium, study their incentive properties, prove dominance relations among them in a simplified setting, and confirm our theoretical findings using real ad auction data from sponsored search. Our study sheds light on the impact of budget management strategies on the tradeoff between the seller's profit and buyers' utility and may be of practical relevance for advertising platforms.

Budget Management Strategies in Repeated Auctions

We present new mechanisms for \emph{label differential privacy}, a relaxation of differentially private machine learning that only protects the privacy of the labels in the training set. Our mechanisms cluster the examples in the training set using their (non-private) feature vectors, randomly re-sample each label from examples in the same cluster, and output a training set with noisy labels as well as a modified version of the true loss function. We prove that when the clusters are both large and high-quality, the model that minimizes the modified loss on the noisy training set converges to small excess risk at a rate that is comparable to the rate for non-private learning. We describe both a centralized mechanism in which the entire training set is stored by a trusted curator, and a distributed mechanism where each user stores a single labeled example and replaces her label with the label of a randomly selected user from the same cluster. We also describe a learning problem in which large clusters are necessary to achieve both strong privacy and either good precision or good recall. Our experiments show that randomizing the labels within each cluster significantly improves the privacy vs. accuracy trade-off compared to applying uniform randomized response to the labels, and also compared to learning a model via DP-SGD.

Label differential privacy via clustering.

Service providers rely on the management systems housed in their Network Operations Centers (NOCs) to remotely operate, monitor and provision their data networks. Lately there has been a tremendous increase in management traffic due to the growing complexity and size of the data networks and the services provisioned on them.

Traffic engineering for management flows is essential for the smooth functioning of these networks to avoid congestion, which can result in loss of critical data such as billing records, network alarms, etc. As is the case with most intra-domain routing protocols, the management flows in many of these networks are routed on shortest paths connecting the NOC with the service provider’s POPs (points of presence). This collection of paths thus forms a “confluent” tree rooted at the gateway router connected to the NOC. The links close to the gateway router may form a bottleneck in this tree resulting in congestion. Typically this congestion is alleviated by adding layer two tunnels (virtual links) that offload the traffic from some links of this tree by routing it directly to the gateway router. The traffic engineering problem is then to minimize the number of virtual links needed for alleviating congestion.

In this paper we formulate a traffic engineering problem motivated by the above mentioned applications. We show that the general versions of this problem are hard to solve. However, for some simpler cases in which the underlying network is a tree, we design efficient algorithms. In particular, we design fully polynomial-time approximate schemes (FPTAS) for different variants of this problem on trees. We use these algorithms as the basis for designing efficient heuristics for alleviating congestion in general (non-tree) service provider network topologies.

/pdf/traffic-engineering-of-management-flows-by-link-39crn6v8do.pdf

Traffic Engineering of Management Flows by Link Augmentations on Confluent Trees

How can we make use of information parallelism in online decision making problems while efficiently balancing the exploration-exploitation trade-off? In this paper, we introduce a batch Thompson Sampling framework for two canonical online decision making problems, namely, stochastic multi-arm bandit and linear contextual bandit with finitely many arms. Over a time horizon $T$, our \textit{batch} Thompson Sampling policy achieves the same (asymptotic) regret bound of a fully sequential one while carrying out only $O(\log T)$ batch queries. To achieve this exponential reduction, i.e., reducing the number of interactions from $T$ to $O(\log T)$, our batch policy dynamically determines the duration of each batch in order to balance the exploration-exploitation trade-off. We also demonstrate experimentally that dynamic batch allocation dramatically outperforms natural baselines such as static batch allocations.

Vahab Mirrokni

Papers

Bandits with adversarial scaling

Budget Management Strategies in Repeated Auctions

Label differential privacy via clustering.

Traffic Engineering of Management Flows by Link Augmentations on Confluent Trees

Parallelizing Thompson Sampling