Proving the performance of a new revenue management system

Revenue management (RM) is a complicated business process that can best be described ascontrol of sales (using prices, restrictions, or capacity), usually using software as a tool to aiddecisions. RM software can play a mere informative role, supplying analysts with formatted andsummarized data who use it to make control decisions (setting a price or allocating capacity fora price point), or, play a deeper role, automating the decisions process completely, at the otherextreme. The RM models and algorithms in the academic literature by and large concentrateon the latter, completely automated, level of functionality.A firm considering using a new RM model or RM system needs to evaluate its performance.Academic papers justify the performance of their models using simulations, where customerbooking requests are simulated according to some process and model, and the revenue perfor-mance of the algorithm compared to an alternate set of algorithms. Such simulations, whilean accepted part of the academic literature, and indeed providing research insight, often lackcredibility with management. Even methodologically, they are usually awed, as the simula-tions only test \within-model" performance, and say nothing as to the appropriateness of themodel in the first place. Even simulations that test against alternate models or competition arelimited by their inherent necessity on fixing some model as the universe for their testing. Theseproblems are exacerbated with RM models that attempt to model customer purchase behav-ior or competition, as the right models for competitive actions or customer purchases remainsomewhat of a mystery, or at least with no consensus on their validity.How then to validate a model? Putting it another way, we want to show that a particularmodel or algorithm is the cause of a certain improvement to the RM process compared to theexisting process. We take care to emphasize that we want to prove the said model as the causeof performance, and to compare against a (incumbent) process rather than against an alternatemodel.In this paper we describe a \live" testing experiment that we conducted at Iberia Airlineson a set of flights. A set of competing algorithms control a set of flights during adjacentweeks, and their behavior and results are observed over a relatively long period of time (9months). In parallel, a group of control flights were managed using the traditional mix of manualand algorithmic control (incumbent system). Such \sandbox" testing, while common at manylarge internet search and e-commerce companies is relatively rare in the revenue managementarea. Sandbox testing has an undisputable model of customer behavior but the experimentaldesign and analysis of results is less clear. In this paper we describe the philosophy behind theexperiment, the organizational challenges, the design and setup of the experiment, and outlinethe analysis of the results. This paper is a complement to a (more technical) related paper thatdescribes the econometrics and statistical analysis of the results.

A simple randomized algorithm for consistent sequential prediction of ergodic time series

We present a simple randomized procedure for the prediction of a binary sequence. The algorithm uses ideas from recent developments of the theory of the prediction of individual sequences. We show that if thesequence is a realization of a stationary and ergodic random process then the average number of mistakes converges, almost surely, to that of the optimum, given by the Bayes predictor.

Worst-case bounds for the logarithmic loss of predictors

We investigate on-line prediction of individual sequences. Given a class of predictors, the goal is to predict as well as the best predictor in the class, where the loss is measured by the self information (logarithmic) loss function. The excess loss (regret) is closely related to the redundancy of the associated lossless universal code. Using Shtarkov's theorem and tools from empirical process theory, we prove a general upper bound on the best possible (minimax) regret. The bound depends on certain metric properties of the class of predictors. We apply the bound to both parametric and nonparametric classes ofpredictors. Finally, we point out a suboptimal behavior of the popular Bayesian weighted average algorithm.

Audit report on the City of Center Point, Iowa for the year ended June 30, 2008

Audit report on the City of Center Point, Iowa for the year ended June 30, 2008

Decision making in uncertain and changing environments

We consider an agent who has to repeatedly make choices in an uncertainand changing environment, who has full information of the past, who discountsfuture payoffs, but who has no prior. We provide a learning algorithm thatperforms almost as well as the best of a given finite number of experts orbenchmark strategies and does so at any point in time, provided the agentis sufficiently patient. The key is to find the appropriate degree of forgettingdistant past. Standard learning algorithms that treat recent and distant pastequally do not have the sequential epsilon optimality property.

Rein et tabac : revue de la littérature et mise au point [Kidney and smoking: literature review and focus].

Smoking remains a major public health problem. It is associated with a considerable number of deaths in the world's population. Smoking is just like high blood pressure, an independent predictor of progression to any primary renal disease and renal transplant patients. It seems that smoking cessation slows the progression of kidney disease in smokers. The literature data are sometimes contradictory about it because of some methodological weaknesses. However, experimental models highlight the harmful effects of tobacco by hemodynamic and non-hemodynamic factors. The conclusion is that a major effort should be further produced by the nephrology community to motivate our patients to stop smoking.

Human immunodeficiency virus type 1 fitness is a determining factor in viral rebound and set point in chronic infection.

Human immunodeficiency virus type 1 (HIV-1) isolates from 20 chronically infected patients who participated in a structured treatment interruption (STI) trial were studied to determine whether viral fitness influences reestablishment of viremia. Viruses derived from individuals who spontaneously controlled viremia had significantly lower in vitro replication capacities than viruses derived from individuals that did not control viremia after interruption of antiretroviral therapy (ART), and replication capacities correlated with pre-ART and post-STI viral set points. Of note, no clinically relevant improvement of viral loads upon STI occurred. Virus isolates from controlling and noncontrolling patients were indistinguishable in terms of coreceptor usage, genetic subtype, and sensitivity to neutralizing antibodies. In contrast, viruses from controlling patients exhibited increased sensitivity to inhibition by chemokines. Sensitivity to inhibition by RANTES correlated strongly with slower replication kinetics of the virus isolates, suggesting a marked dependency of these virus isolates on high coreceptor densities on the target cells. In summary, our data indicate that viral fitness is a driving factor in determining the magnitude of viral rebound and viral set point in chronic HIV-1 infection, and thus fitness should be considered as a parameter influencing the outcome of therapeutic intervention in chronic infection.

Mise au point et validation d'un fantôme spécifique pour asssurer le suivi des mesures quantitatives en tomographie par émission de positons.

L'index de consommation du glucose, SUV pour standardized uptake value, est largement utilisé en tomographie par émission de positons (TEP) pour quantifier la concentration relative de [18F]2-fluoro-2-désoxy-D-glucose (18F-FDG) dans les tissus. Cependant, cet indice dépend de nombreux facteurs méthodologiques et biologiques et son utilisation fait débat. Il est donc primordial d'instaurer un contrôle qualité régulier permettant d'assurer la stabilité de la mesure des indices quantitatifs. Dans cette optique, un fantôme spécifique avec inserts cylindriques de différentes tailles a été développé. L'objectif de cette étude est de montrer la sensibilité et l'intérêt de ce fantôme. Méthodes. - La sensibilité du fantôme a été étudiée à travers la mesure des SUV et des coefficients de recouvrement (RC). Plusieurs méthodes de mesure ont été utilisées. Les données ont été reconstruites en utilisant les algorithmes de routine clinique. Nous avons étudié la variation des RC en fonction de la taille des cylindres et le changement relatif de fixation, en utilisant des activités différentes. Le fantôme a ensuite été testé sur l'appareil d'un autre centre. Résultats. - Pour toutes les méthodes de mesure, une forte variation des RC avec la taille des cylindres, de l'ordre de 50 %, a été obtenue. Ce fantôme a également permis de mesurer un changement relatif de fixation qui s'est révélé être indépendant de la méthode de mesure. Malgré un étalonnage des deux systèmes TEP/TDM, des différences de quantification d'environ 20 % ont subsisté. Conclusion. - Les résultats obtenus montrent l'intérêt de ce fantôme dans le cadre du suivi des mesures quantitatives.

Restoration of a lost geological reference point: the loggingstone of Pedralta (Girona, Spain)

The Pedralta is a loggingstone formed of monzonitic leucogranites, which has a volume of about 38 m3 and weighs 101 tons. After its fall in December 1996, a public subscription was opened to meet the expenses of the restoration work, which was completed in May 1999