Software and hardware co-design and optimization of HPC systems has become intolerably complex, ad-hoc, time consuming and error prone due to enormous number of available design and optimization choices, complex interactions between all software and hardware components, and multiple strict requirements placed on performance, power consumption, size, reliability and cost. We present our novel long-term holistic and practical solution to this problem based on customizable, plugin-based, schema-free, heterogeneous, open-source Collective Mind repository and infrastructure with unified web interfaces and on-line advise system. This collaborative framework distributes analysis and multi-objective off-line and on-line auto-tuning of computer systems among many participants while utilizing any available smart phone, tablet, laptop, cluster or data center, and continuously observing, classifying and modeling their realistic behavior. Any unexpected behavior is analyzed using shared data mining and predictive modeling plugins or exposed to the community at cTuning.org for collaborative explanation, top-down complexity reduction, incremental problem decomposition and detection of correlating program, architecture or run-time properties (features). Gradually increasing optimization knowledge helps to continuously improve optimization heuristics of any compiler, predict optimizations for new programs or suggest efficient run-time (online) tuning and adaptation strategies depending on end-user requirements. We decided to share all our past research artifacts including hundreds of codelets, numerical applications, data sets, models, universal experimental analysis and auto-tuning pipelines, self-tuning machine learning based meta compiler, and unified statistical analysis and machine learning plugins in a public repository to initiate systematic, reproducible and collaborative research, development and experimentation with a new publication model where experiments and techniques are validated, ranked and improved by the community.

Linear active-power-only DC power flow approximations are pervasive in the planning and control of power systems. However, these approximations fail to capture reactive power and voltage magnitudes, both of which are necessary in many applications to ensure voltage stability and AC power flow feasibility. This paper proposes linear-programming models (the LPAC models) that incorporate reactive power and voltage magnitudes in a linear power flow approximation. The LPAC models are built on a convex approximation of the cosine terms in the AC equations, as well as Taylor approximations of the remaining nonlinear terms. Experimental comparisons with AC solutions on a variety of standard IEEE and MatPower benchmarks show that the LPAC models produce accurate values for active and reactive power, phase angles, and voltage magnitudes. The potential benefits of the LPAC models are illustrated on two "proof-of-concept" studies in power restoration and capacitor placement.

The iPDB procedure by Haslum et al. is the state-of-the-art method for computing additive abstraction heuristics for domain-independent planning. It performs a hill-climbing search in the space of pattern collections, combining information from multiple patterns in the so-called canonical heuristic. We show how stronger heuristic estimates can be obtained through linear programming. An experimental evaluation demonstrates the strength of the new technique on the IPC benchmark suite.

The semi-supervised learning usually only predict labels for unlabeled data appearing in training data, and cannot effectively predict labels for testing data never appearing in training set. To handle this out-of-sample problem, many inductive methods make a constraint such that the predicted label matrix should be exactly equal to a linear model. In practice, this constraint is too rigid to capture the manifold structure of data. Motivated by this deficiency, we relax the rigid linear embedding constraint and propose to use an elastic embedding constraint on the predicted label matrix such that the manifold structure can be better explored. To solve our new objective and also a more general optimization problem, we study a novel adaptive loss with efficient optimization algorithm. Our new adaptive loss minimization method takes the advantages of both L1 norm and L2 norm, and is robust to the data outlier under Laplacian distribution and can efficiently learn the normal data under Gaussian distribution. Experiments have been performed on image classification tasks and our approach outperforms other state-of-the-art methods.

Evolutionary algorithms (EAs) are a large family of heuristic optimization algorithms inspired by natural phenomena, and are often used in practice to obtain satisficing instead of optimal solutions. In this work, we investigate a largely underexplored issue: the approximation performance of EAs in terms of how close the obtained solution is to an optimal solution. We study an EA framework named simple EA with isolated population (SEIP) that can be implemented as a single- or multi-objective EA. We present general approximation results of SEIP, and specifically on the minimum set cover problem, we find that SEIP achieves the currently best-achievable approximation ratio. Moreover, on an instance class of the k -set cover problem, we disclose how SEIP can overcome the difficulty that limits the greedy algorithm.

Diffusion processes in networks are increasingly used to model dynamic phenomena such as the spread of information, wildlife, or social influence. Our work addresses the problem of learning the underlying parameters that govern such a diffusion process by observing the time at which nodes become active. A key advantage of our approach is that, unlike previous work, it can tolerate missing observations for some nodes in the diffusion process. Having incomplete observations is characteristic of offline networks used to model the spread of wildlife. We develop an EM algorithm to address parameter learning in such settings. Since both the E and M steps are computationally challenging, we employ a number of optimization methods such as nonlinear and difference-of-convex programming to address these challenges. Evaluation of the approach on the Red-cockaded Woodpecker conservation problem shows that it is highly robust and accurately learns parameters in various settings, even with more than 80% missing data.

Constraint-based applications such as configurators, recommenders, and scheduling systems support users in complex decision making scenarios. Typically, these systems try to identify a solution that satisfies all articulated user requirements. If the requirements are inconsistent with the underlying constraint set, users have to be actively supported in finding a way out from the no solution could be found dilemma. In this paper we introduce techniques that support the calculation of personalized diagnoses for inconsistent constraint sets. These techniques significantly improve the diagnosis prediction quality compared to approaches based on the calculation of minimal cardinality diagnoses. In order to show the applicability of our approach we present the results of an empirical study and a corresponding performance analysis.

In this paper we study a dynamic problem of ridesharing and taxi sharing with time windows. We consider a scenario where people needing a taxi or interested in getting a ride use a phone app to designate their source and destination points in a city, as well others restrictions (such as maximum allowable time to be at the destination). On the other hand, we have taxis and people interested in giving a ride, with their current positions and also some constraints (vehicle capacity, destination, maximum time to destination). We want to maximize the number of shared trips: in the case of taxis, people going to close locations can share the costs of the trip, and in case of rides, the driver and passengers can share costs as well. This problem is dynamic since new calls for taxis or calls for rides arrive on demand. This give rise to an optimization problem which we prove to be NP-Hard. We then propose heuristics to deal with it. We focus on the taxi sharing problem, but we show that our model is easily extendable to model the ridesharing situation or even a situation where there are both taxis and car owners. In addition, we present a framework that consists basically of a client aplication and a server. The last one processes all incoming information in order to match vehicles to passengers requests. The entire system can be used by taxi companies and riders in a way to reduce traffic in the cities and to reduce the emission of greenhouse gases.

Multi-label classification is a critical problem in many areas of data analysis such as image labeling and text categorization. In this paper we propose a probabilistic multi-label classification model based on novel sparse feature learning. By employing an individual sparsity inducing ℓ1-norm and a group sparsity inducing ℓ2,1-norm, the proposed model has the capacity of capturing both label interdependencies and common predictive model structures. We formulate this sparse norm regularized learning problem as a non-smooth convex optimization problem, and develop a fast proximal gradient algorithm to solve it for an optimal solution. Our empirical study demonstrates the efficacy of the proposed method on a set of multi-label tasks given a limited number of labeled training instances.

Bayesian optimization techniques have been successfully applied to robotics, planning, sensor placement, recommendation, advertising, intelligent user interfaces and automatic algorithm configuration. Despite these successes, the approach is restricted to problems of moderate dimension, and several workshops on Bayesian optimization have identified its scaling to high dimensions as one of the holy grails of the field. In this paper, we introduce a novel random embedding idea to attack this problem. The resulting Random EMbedding Bayesian Optimization (REMBO) algorithm is very simple and applies to domains with both categorical and continuous variables. The experiments demonstrate that REMBO can effectively solve high-dimensional problems, including automatic parameter configuration of a popular mixed integer linear programming solver.