Learning for scheduling a portfolio of constraint solvers

Nel lavoro di tesi qui presentato si indaga l'applicazione di tecniche di apprendimento mirate ad una più efficiente esecuzione di un portfolio di risolutore di vincoli (constraint solver). Un constraint solver è un programma che dato in input un problema di vincoli, elabora una soluzione mediante l'utilizzo di svariate tecniche. I problemi di vincoli sono altamente presenti nella vita reale. Esempi come l'organizzazione dei viaggi dei treni oppure la programmazione degli equipaggi di una compagnia aerea, sono tutti problemi di vincoli. Un problema di vincoli è formalizzato da un problema di soddisfacimento di vincoli(CSP). Un CSP è descritto da un insieme di variabili che possono assumere valori appartenenti ad uno specico dominio ed un insieme di vincoli che mettono in relazione variabili e valori assumibili da esse. Una tecnica per ottimizzare la risoluzione di tali problemi è quella suggerita da un approccio a portfolio. Tale tecnica, usata anche in am- biti come quelli economici, prevede la combinazione di più solver i quali assieme possono generare risultati migliori di un approccio a singolo solver. In questo lavoro ci preoccupiamo di creare una nuova tecnica che combina un portfolio di constraint solver con tecniche di machine learning. Il machine learning è un campo di intelligenza articiale che si pone l'obiettivo di immettere nelle macchine una sorta di `intelligenza'. Un esempio applicativo potrebbe essere quello di valutare i casi passati di un problema ed usarli in futuro per fare scelte. Tale processo è riscontrato anche a livello cognitivo umano. Nello specico, vogliamo ragionare in termini di classicazione. Una classicazione corrisponde ad assegnare ad un insieme di caratteristiche in input, un valore discreto in output, come vero o falso se una mail è classicata come spam o meno. La fase di apprendimento sarà svolta utilizzando una parte di CPHydra, un portfolio di constraint solver sviluppato presso la University College of Cork (UCC). Di tale algoritmo a portfolio verranno utilizzate solamente le caratteristiche usate per descrivere determinati aspetti di un CSP rispetto ad un altro; queste caratteristiche vengono altresì dette features. Creeremo quindi una serie di classicatori basati sullo specifico comportamento dei solver. La combinazione di tali classicatori con l'approccio a portfolio sara nalizzata allo scopo di valutare che le feature di CPHydra siano buone e che i classicatori basati su tali feature siano affidabili. Per giusticare il primo risultato, eettueremo un confronto con uno dei migliori portfolio allo stato dell'arte, SATzilla. Una volta stabilita la bontà delle features utilizzate per le classicazioni, andremo a risolvere i problemi simulando uno scheduler. Tali simulazioni testeranno diverse regole costruite con classicatori precedentemente introdotti. Prima agiremo su uno scenario ad un processore e successivamente ci espanderemo ad uno scenario multi processore. In questi esperimenti andremo a vericare che, le prestazioni ottenute tramite l'applicazione delle regole create appositamente sui classicatori, abbiano risultati migliori rispetto ad un'esecuzione limitata all'utilizzo del migliore solver del portfolio. I lavoro di tesi è stato svolto in collaborazione con il centro di ricerca 4C presso University College Cork. Su questo lavoro è stato elaborato e sottomesso un articolo scientico alla International Joint Conference of Articial Intelligence (IJCAI) 2011. Al momento della consegna della tesi non siamo ancora stati informati dell'accettazione di tale articolo. Comunque, le risposte dei revisori hanno indicato che tale metodo presentato risulta interessante.

Constraint based methods for allocation and scheduling of periodic applications

This work presents exact algorithms for the Resource Allocation and Cyclic Scheduling Problems (RA&CSPs). Cyclic Scheduling Problems arise in a number of application areas, such as in hoist scheduling, mass production, compiler design (implementing scheduling loops on parallel architectures), software pipelining, and in embedded system design. The RA&CS problem concerns time and resource assignment to a set of activities, to be indefinitely repeated, subject to precedence and resource capacity constraints. In this work we present two constraint programming frameworks facing two different types of cyclic problems. In first instance, we consider the disjunctive RA&CSP, where the allocation problem considers unary resources. Instances are described through the Synchronous Data-flow (SDF) Model of Computation. The key problem of finding a maximum-throughput allocation and scheduling of Synchronous Data-Flow graphs onto a multi-core architecture is NP-hard and has been traditionally solved by means of heuristic (incomplete) algorithms. We propose an exact (complete) algorithm for the computation of a maximum-throughput mapping of applications specified as SDFG onto multi-core architectures. Results show that the approach can handle realistic instances in terms of size and complexity. Next, we tackle the Cyclic Resource-Constrained Scheduling Problem (i.e. CRCSP). We propose a Constraint Programming approach based on modular arithmetic: in particular, we introduce a modular precedence constraint and a global cumulative constraint along with their filtering algorithms. Many traditional approaches to cyclic scheduling operate by fixing the period value and then solving a linear problem in a generate-and-test fashion. Conversely, our technique is based on a non-linear model and tackles the problem as a whole: the period value is inferred from the scheduling decisions. The proposed approaches have been tested on a number of non-trivial synthetic instances and on a set of realistic industrial instances achieving good results on practical size problem.

Performance evaluation of improved output scheduling in 3lihon nodes

Le reti ottiche, grazie alla loro elevata capacità, hanno acquisito sempre maggiore importanza negli ultimi anni, sia per via del crescente volume di dati scambiati, legato soprattutto alla larga diffusione di Internet, sia per la necessità di comunicazioni in tempo reale. Dati i (relativamente) lunghi tempi di adattamento, questa tecnologia nativamente non è quella ottimale per il trasporto di un traffico a burst, tipico delle telecomunicazioni odierne. Le reti ibride cercano, quindi, di coniugare al meglio le potenzialità della commutazione ottica di circuito e della commutazione ottica a pacchetto. In questo lavoro, in particolare, ci si è concentrati su un'architettura di rete ibrida denominata 3LIHON (3-Level Integrated Hybrid Optical Network). Essa prevede tre distinti livelli di qualità di servizio (QoS) per soddisfare differenti necessità: - Guaranteed Service Type (GST): simile ad un servizio a commutazione di circuito, non ammette perdita di dati. - Statistically Multiplexed Real Time (SM/RT): simile ad un servizio a commutazione di pacchetto, garantisce ritardo nullo o molto basso all'interno della rete, permette un piccolo tasso di perdita di dati e ammette la contesa della banda. - Statistically Multiplexed Best Effort (SM/BE): simile ad un servizio a commutazione di pacchetto, non garantisce alcun ritardo tra i nodi ed ammette un basso tasso di perdita dei dati. In un nodo 3LIHON, il traffico SM/BE impossibile da servire, a causa ad es. dell'interruzione da parte di pacchetti aventi un livello di QoS più prioritario, viene irrimediabilmente perso. Questo implica anche lo spreco del tempo e delle risorse impiegati per trasmettere un pacchetto SM/BE fino alla sua interruzione. Nel presente lavoro si è cercato di limitare, per quanto possibile, questo comportamento sconveniente, adottando e comparando tre strategie, che hanno portato alla modifica del nodo 3LIHON standard ed alla nascita di tre sue varianti.

Optimized scheduling in real-time environments with column generation

Im Bereich sicherheitsrelevanter eingebetteter Systeme stellt sich der Designprozess von Anwendungen als sehr komplex dar. Entsprechend einer gegebenen Hardwarearchitektur lassen sich Steuergeräte aufrüsten, um alle bestehenden Prozesse und Signale pünktlich auszuführen. Die zeitlichen Anforderungen sind strikt und müssen in jeder periodischen Wiederkehr der Prozesse erfüllt sein, da die Sicherstellung der parallelen Ausführung von größter Bedeutung ist. Existierende Ansätze können schnell Designalternativen berechnen, aber sie gewährleisten nicht, dass die Kosten für die nötigen Hardwareänderungen minimal sind. Wir stellen einen Ansatz vor, der kostenminimale Lösungen für das Problem berechnet, die alle zeitlichen Bedingungen erfüllen. Unser Algorithmus verwendet Lineare Programmierung mit Spaltengenerierung, eingebettet in eine Baumstruktur, um untere und obere Schranken während des Optimierungsprozesses bereitzustellen. Die komplexen Randbedingungen zur Gewährleistung der periodischen Ausführung verlagern sich durch eine Zerlegung des Hauptproblems in unabhängige Unterprobleme, die als ganzzahlige lineare Programme formuliert sind. Sowohl die Analysen zur Prozessausführung als auch die Methoden zur Signalübertragung werden untersucht und linearisierte Darstellungen angegeben. Des Weiteren präsentieren wir eine neue Formulierung für die Ausführung mit fixierten Prioritäten, die zusätzlich Prozessantwortzeiten im schlimmsten anzunehmenden Fall berechnet, welche für Szenarien nötig sind, in denen zeitliche Bedingungen an Teilmengen von Prozessen und Signalen gegeben sind. Wir weisen die Anwendbarkeit unserer Methoden durch die Analyse von Instanzen nach, welche Prozessstrukturen aus realen Anwendungen enthalten. Unsere Ergebnisse zeigen, dass untere Schranken schnell berechnet werden können, um die Optimalität von heuristischen Lösungen zu beweisen. Wenn wir optimale Lösungen mit Antwortzeiten liefern, stellt sich unsere neue Formulierung in der Laufzeitanalyse vorteilhaft gegenüber anderen Ansätzen dar. Die besten Resultate werden mit einem hybriden Ansatz erzielt, der heuristische Startlösungen, eine Vorverarbeitung und eine heuristische mit einer kurzen nachfolgenden exakten Berechnungsphase verbindet.

Scheduling di applicazioni periodiche con durata variabile

In questa tesi presentiamo una strategia, e la relativa implementazione, per il problema dell’allocazione e schedulazione, su risorse unarie, di applicazioni multi-task periodiche, composte da attività che interagiscono fra loro e la cui durata è incerta. Lo scopo che ci si propone di raggiungere, è l’implementazione di una strategia di allocazione schedulazione che garantisca robustezza ed efficienza, in quei contesti in cui la conoscenza a priori è limitata e in cui le applicazioni si ripetono indefinitamente nel tempo. Per raggiungere questo scopo, sarà usato un approccio ibrido fra statico e dinamico. Staticamente è generata una soluzione del problema, sfruttando la programmazione a vincoli, in cui le durate delle attività sono arbitrariamente fissate. Questa soluzione, non rappresenta la soluzione del nostro problema, ma è utilizzata per generare un ordinamento delle attività, che compongono le applicazioni periodiche. Dinamicamente, sfruttando l’ordinamento ottenuto, è effettuata l’allocazione e la schedulazione effettiva delle applicazioni periodiche, considerando durate variabili per le attività. L’efficienza ottenuta applicando il nostro approccio è valutata effettuando test su una vasta gamma di istanze, sia industriali, sia sintetiche appositamente generate. I risultati sono confrontati con quelli ottenuti, per le stesse istanze, applicando un approccio puramente statico. Come si vedrà, in alcuni casi, è possibile anche quadruplicale la velocità di completamento delle applicazioni trattate.

Distributed allocation and scheduling of applications on HPC platforms

High Performance Computing e una tecnologia usata dai cluster computazionali per creare sistemi di elaborazione che sono in grado di fornire servizi molto piu potenti rispetto ai computer tradizionali. Di conseguenza la tecnologia HPC e diventata un fattore determinante nella competizione industriale e nella ricerca. I sistemi HPC continuano a crescere in termini di nodi e core. Le previsioni indicano che il numero dei nodi arrivera a un milione a breve. Questo tipo di architettura presenta anche dei costi molto alti in termini del consumo delle risorse, che diventano insostenibili per il mercato industriale. Un scheduler centralizzato non e in grado di gestire un numero di risorse cosi alto, mantenendo un tempo di risposta ragionevole. In questa tesi viene presentato un modello di scheduling distribuito che si basa sulla programmazione a vincoli e che modella il problema dello scheduling grazie a una serie di vincoli temporali e vincoli sulle risorse che devono essere soddisfatti. Lo scheduler cerca di ottimizzare le performance delle risorse e tende ad avvicinarsi a un profilo di consumo desiderato, considerato ottimale. Vengono analizzati vari modelli diversi e ognuno di questi viene testato in vari ambienti.

MOWER : A NEW DESIGN FOR NON-BLOCKING MISPREDICTION RECOVERY

Mower is a micro-architecture technique which targets branch misprediction penalties in superscalar processors. It speeds-up the misprediction recovery process by dynamically evicting stale instructions and fixing the RAT (Register Alias Table) using explicit branch dependency tracking. Tracking branch dependencies is accomplished by using simple bit matrices. This low-overhead technique allows overlapping of the recovery process with instruction fetching, renaming and scheduling from the correct path. Our evaluation of the mechanism indicates that it yields performance very close to ideal recovery and provides up to 5% speed-up and 2% reduction in power consumption compared to a traditional recovery mechanism using a reorder buffer and a walker. The simplicity of the mechanism should permit easy implementation of Mower in an actual processor.

A hybrid method for large-scale short-term scheduling of make-and-pack production processes

Due to the ongoing trend towards increased product variety, fast-moving consumer goods such as food and beverages, pharmaceuticals, and chemicals are typically manufactured through so-called make-and-pack processes. These processes consist of a make stage, a pack stage, and intermediate storage facilities that decouple these two stages. In operations scheduling, complex technological constraints must be considered, e.g., non-identical parallel processing units, sequence-dependent changeovers, batch splitting, no-wait restrictions, material transfer times, minimum storage times, and finite storage capacity. The short-term scheduling problem is to compute a production schedule such that a given demand for products is fulfilled, all technological constraints are met, and the production makespan is minimised. A production schedule typically comprises 500–1500 operations. Due to the problem size and complexity of the technological constraints, the performance of known mixed-integer linear programming (MILP) formulations and heuristic approaches is often insufficient. We present a hybrid method consisting of three phases. First, the set of operations is divided into several subsets. Second, these subsets are iteratively scheduled using a generic and flexible MILP formulation. Third, a novel critical path-based improvement procedure is applied to the resulting schedule. We develop several strategies for the integration of the MILP model into this heuristic framework. Using these strategies, high-quality feasible solutions to large-scale instances can be obtained within reasonable CPU times using standard optimisation software. We have applied the proposed hybrid method to a set of industrial problem instances and found that the method outperforms state-of-the-art methods.

On an Application of Microsoft Excel’s Evolutionary Solver to the Resource-Constrained Project Scheduling Problem RCPSP

The Solver Add-in of Microsoft Excel is widely used in courses on Operations Research and in industrial applications. Since the 2010 version of Microsoft Excel, the Solver Add-in comprises a so-called evolutionary solver. We analyze how this metaheuristic can be applied to the resource-constrained project scheduling problem (RCPSP). We present an implementation of a schedule-generation scheme in a spreadsheet, which combined with the evolutionary solver can be used for devising good feasible schedules. Our computational results indicate that using this approach, non-trivial instances of the RCPSP can be (approximately) solved to optimality.

An MILP-based heuristic for staff scheduling problems with acceptance levels

We present a real-world staff-assignment problem that was reported to us by a provider of an online workforce scheduling software. The problem consists of assigning employees to work shifts subject to a large variety of requirements related to work laws, work shift compatibility, workload balancing, and personal preferences of employees. A target value is given for each requirement, and all possible deviations from these values are associated with acceptance levels. The objective is to minimize the total number of deviations in ascending order of the acceptance levels. We present an exact lexicographic goal programming MILP formulation and an MILP-based heuristic. The heuristic consists of two phases: in the first phase a feasible schedule is built and in the second phase parts of the schedule are iteratively re-optimized by applying an exact MILP model. A major advantage of such MILP-based approaches is the flexibility to account for additional constraints or modified planning objectives, which is important as the requirements may vary depending on the company or planning period. The applicability of the heuristic is demonstrated for a test set derived from real-world data. Our computational results indicate that the heuristic is able to devise optimal solutions to non-trivial problem instances, and outperforms the exact lexicographic goal programming formulation on medium- and large-sized problem instances.

4HPR-X radiation mechanisms of interactions in non-small cell lung cancer (NSCLC) cells in vitro

Lung cancer is the leading cause of cancer death. However, poor survival using conventional therapies fuel the search for more rational interventions. The objective of this study was to design and implement a 4HPR-radiation interaction model in NSCLC, employing a traditional clinical modality (radiation), a relatively new, therapeutically unexplored agent (4HPR) and rationally combining them based on molecular mechanistic findings pertaining to their interactions. To test the hypothesis that 4HPR sensitizes cells to radiation-induced cell death via G2+M accumulation, we designed a working model consisting of H522 adenocarcinoma cells (p53, K-ras mutated) derived from an NSCLC patient; 4HPR at concentrations up to 10 μM; and X radiation up to 6 Gy generated by a patient-dedicated Phillips RT-250 X ray unit at 250 KV, 15 mA, 1.85 Gy/min. We found that 4HPR produced time- and dose-dependent morphological changes, growth inhibition, and DNA damage-inducing enhancement of reactive oxygen species. A transient G2+M accumulation of cells maximal at 24 h of continuous 4HPR exposure was used for irradiation time scheduling. Our data demonstrated enhanced cell death (both apoptotic and necrotic) in irradiated cells pre-treated with 4HPR versus those with either stressor alone. 4HPR's effect of increased NSCLC cells' radioresponse was confirmed by clonogenic assay. To explore these practical findings from a molecular mechanistic perspective, we further investigated and showed that levels of cyclin B1 and p34cdc2 kinase—both components of the mitosis promoting factor (MPF) regulating the G2/M transition—did not change following 4HPR treatment. Likewise, cdc25C phosphatase was not altered. However, enhanced p34cdc2 phosphorylation on its Thr14Tyr15 residues—indicative of its inactivation and increased expression of MPF negative regulators chk1 and wee1 kinases—were supportive of explaining 4HPR-treated cells' accumulation. Hence, p34cdc2 phosphorylation, chk1, and wee1 warrant further evaluation as potential molecular targets for 4HPR-X radiation combination. In summary, we (1) demonstrated that 4HPR not only induces cell death by itself, but also increases NSCLC cells' subsequent radioresponse, indicative of potential clinical applicability, and (2) for the first time, shed light on deciphering 4HPR-X radiation molecular mechanisms of interaction, including the finding of 4HPR's role as a p34cdc2 inactivator via Thr14Tyr15 phosphorylation. ^

Relating goal scheduling, precedence, and memory management in and-parallel execution of logic programs

The interactions among three important issues involved in the implementation of logic programs in parallel (goal scheduling, precedence, and memory management) are discussed. A simplified, parallel memory management model and an efficient, load-balancing goal scheduling strategy are presented. It is shown how, for systems which support "don't know" non-determinism, special care has to be taken during goal scheduling if the space recovery characteristics of sequential systems are to be preserved. A solution based on selecting only "newer" goals for execution is described, and an algorithm is proposed for efficiently maintaining and determining precedence relationships and variable ages across parallel goals. It is argued that the proposed schemes and algorithms make it possible to extend the storage performance of sequential systems to parallel execution without the considerable overhead previously associated with it. The results are applicable to a wide class of parallel and coroutining systems, and they represent an efficient alternative to "all heap" or "spaghetti stack" allocation models.

Purchasing behaviour for non-prescription medicines

Background: The non-prescription medicine, market is constantly challenges. With changes to scheduling and market dynamics, a need for current Australian data on medicines purchasing behaviour was identified. Objectives: This survey aimed to report on the purchasing behaviour of non-prescription medicine customers, the medicines bought and influences on medicine sales. Methods: Researchers were stationed in 15 community pharmacies in southeast Queensland during mid-August 2004. Interview and observational data were collected for all eligible medicine purchases -over approximately 35 hours per pharmacy. Results: Data were collected for 3017 medicines purchased by 2583 customers. Most purchases were made by females (65%) and customers aged 26-35 years (25.8%). Pharmacy assistants alone provided advice in 58% of sales. Two thirds of purchases were for self use. In two thirds of cases, customers had a particular brand in mind; this was highly correlated with previous purchases. Pharmacy staff were highly influential in first time purchases. Conclusions: This study reports a high level of involvement and influence of pharmacy staff in medicine selection.

The effect of non-standard employment on work-family balance

Across the last four decades, the structure of the Australian labour market has changed profoundly as non-standard forms of employment have become more prevalent. According to many researchers, the growth of non-standard work has been driven by employee preferences, particularly among married women, for greater flexibility to balance paid work with domestic responsibilities and other non-work related pursuits. In contrast, other researchers argue that the increasing prevalence of non-standard employment reflects employer demands for greater staffing flexibility. From this perspective, non-standard forms of employment are considered to have a negative effect on work-family balance. This paper explores whether non-standard employment is associated with improved or poorer work-to-family conflict and tests whether experiences vary by gender. It concentrates on three common forms of non-standard employment: part-time employment, casual and fixed-term work contracts and flexible scheduling practices (such as evening work, weekend work and irregular rostering). Analysis is based on 2299 employed parents from the first wave of the Household, Income and Labour Dynamics on Australia (HILDA) project. Results show that few scheduling measures are significant determinants of work-family balance. However, part-time employment is associated with reduced work-to-family strain for both men and women, even after controlling for various other employment and household related characteristics. Casual employment, in contrast, incurs the cost of poorer work-family balance for men. Surprisingly, HILDA data show that overall men experience greater work-to-family strain than women.