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

Most current banner advertising is sold through negotiation thereby incurring large transaction costs and possibly sub-optimal allocations. We propose a new automated system for selling banner advertising. In this system, each advertiser specifies a collection of host webpages which are relevant to his product, a desired total quantity of impressions on these pages, and a maximum per-impression price. The system selects a subset of advertisers as winners and maps each winner to a set of impressions on pages within his desired collection. The distinguishing feature of our system as opposed to current combinatorial allocation mechanisms is that, mimicking the current negotiation system, we guarantee that winners receive at least as many advertising opportunities as they requested or else receive ample compensation in the form of a monetary payment by the host. Such guarantees are essential in markets like banner advertising where a major goal of the advertising campaign is developing brand recognition.

As we show, the problem of selecting a feasible subset of advertisers with maximum total value is inapproximable. We thus present two greedy heuristics and discuss theoretical techniques to measure their performances. Our first algorithm iteratively selects advertisers and corresponding sets of impressions which contribute maximum marginal per-impression profit to the current solution. We prove a bi-criteria approximation for this algorithm, showing that it generates approximately as much value as the optimum algorithm on a slightly harder problem. However, this algorithm might perform poorly on instances in which the value of the optimum solution is quite large, a clearly undesirable failure mode. Hence, we present an adaptive greedy algorithm which again iteratively selects advertisers with maximum marginal per-impression profit, but additionally reassigns impressions at each iteration. For this algorithm, we prove a structural approximation result, a newly defined framework for evaluating heuristics. We thereby prove that this algorithm has a better performance guarantee than the simple greedy algorithm.

/pdf/a-combinatorial-allocation-mechanism-for-banner-okymff210j.pdf

A Combinatorial Allocation Mechanism for Banner Advertisement with Penalties

Motivated by online advertising applications, we study incentive properties of stationary mechanisms that satisfy budget constraints in expectation at a stationary equilibrium. We consider a general repeated auction setting where a seller sells identical items to buyers with budget constraints and the buyers' value distributions can be arbitrarily correlated. We introduce the novel notion of budget-constrained incentive compatibility (BCIC) under which each buyer chooses an optimal bidding strategy among stationary budget-feasible bidding strategies. Armed with the notion of BCIC, we characterize Bayesian optimal mechanisms that satisfy the budget constraints in expectation with respect to the profit, utility and welfare objectives in the restricted setting where the buyers' value distributions are independent. Furthermore, in the general setting where the buyers' value distributions are correlated, we provide the first systematic study on the incentive properties of different budget management mechanisms, including those studied in the literature and, to the best of our knowledge, used in the industry. We explore the following mechanisms and show some popular mechanisms are not incentive compatible even when restricting attention to stationary budget-feasible deviations: throttling, thresholding, bid shading, reserve pricing and two versions of multiplicative boosting.

Budget-Constrained Incentive Compatibility for Stationary Mechanisms

Dynamic mechanisms are a powerful technique in designing revenue-maximizing repeated auctions. Despite their strength, these types of mechanisms have not been widely adopted in practice for several reasons, e.g., for their complexity, and for their sensitivity to the accuracy of predicting buyers» value distributions. In this paper, we aim to address these shortcomings and develop simple dynamic mechanisms that can be implemented efficiently, and provide theoretical guidelines for decreasing the sensitivity of dynamic mechanisms on prediction accuracy of buyers» value distributions. We prove that the dynamic mechanism we propose is provably dynamic incentive compatible, and introduce a notion of buyers» regret in dynamic mechanisms, and show that our mechanism achieves bounded regret while improving revenue and social welfare compared to a static reserve pricing policy. Finally, we confirm our theoretical analysis via an extensive empirical study of our dynamic auction on real data sets from online adverting. For example, we show our dynamic mechanisms can provide a +17% revenue lift with relative regret less than 0.2%.

Dynamic Mechanism Design in the Field

In digital online advertising, advertisers procure ad impressions simultaneously on multiple platforms, or so-called channels, such as Google Ads, Meta Ads Manager, etc., each of which consists of numerous ad auctions. We study how an advertiser maximizes total conversion (e.g. ad clicks) while satisfying aggregate return-on-investment (ROI) and budget constraints across all channels. In practice, an advertiser does not have control over, and thus cannot globally optimize, which individual ad auctions she participates in for each channel, and instead authorizes a channel to procure impressions on her behalf: the advertiser can only utilize two levers on each channel, namely setting a per-channel budget and per-channel target ROI. In this work, we first analyze the effectiveness of each of these levers for solving the advertiser's global multi-channel problem. We show that when an advertiser only optimizes over per-channel ROIs, her total conversion can be arbitrarily worse than what she could have obtained in the global problem. Further, we show that the advertiser can achieve the global optimal conversion when she only optimizes over per-channel budgets. In light of this finding, under a bandit feedback setting that mimics real-world scenarios where advertisers have limited information on ad auctions in each channels and how channels procure ads, we present an efficient learning algorithm that produces per-channel budgets whose resulting conversion approximates that of the global optimal problem. Finally, we argue that all our results hold for both single-item and multi-item auctions from which channels procure impressions on advertisers' behalf.

Multi-channel Autobidding with Budget and ROI Constraints

A major challenge faced by marketers attempting to optimize their advertising campaigns is to deal with budget constraints. The problem is even harder in the face of multidimensional budget constra...

Vahab Mirrokni

Papers

A Combinatorial Allocation Mechanism for Banner Advertisement with Penalties

Budget-Constrained Incentive Compatibility for Stationary Mechanisms

Dynamic Mechanism Design in the Field

Multi-channel Autobidding with Budget and ROI Constraints

Concise Bid Optimization Strategies with Multiple Budget Constraints