Evolving decision trees with beam search-based initialization and lexicographic multi-objective evaluation

Decision tree induction algorithms represent one of the most popular techniques for dealing with classification problems. However, traditional decision-tree induction algorithms implement a greedy approach for node splitting that is inherently susceptible to local optima convergence. Evolutionary algorithms can avoid the problems associated with a greedy search and have been successfully employed to the induction of decision trees. Previously, we proposed a lexicographic multi-objective genetic algorithm for decision-tree induction, named LEGAL-Tree. In this work, we propose extending this approach substantially, particularly w.r.t. two important evolutionary aspects: the initialization of the population and the fitness function. We carry out a comprehensive set of experiments to validate our extended algorithm. The experimental results suggest that it is able to outperform both traditional algorithms for decision-tree induction and another evolutionary algorithm in a variety of application domains.

Studio Sperimentale e Modellazione della Separazione di Proteine con Membrane di Affinità

Chromatography represents one of the most important and widely used unit operation in the biotechnology industry. However this technique suffers from several limitations such as high pressure drop, slow mass transfer through the diffusive pores and strong dependence of the binding capacity on flow rate. In this work, affinity membranes with improved capacity have been considered as an alternative technology for the capturing step in antibody manufacturing. Several affinity membranes have been prepared starting from various membrane supports. Different affinity ligands have been utilized like Protein A, the natural ligand of choice for antibodies, as well as synthetic ligands that exhibit affinity for the Fc portion of antibodies. The membranes have been characterized in detail: binding and elution performance was evaluated in adsorption experiments using pure IgG solutions, while membrane selectivity was evaluated using complex solutions like a cell culture supernatant. The most promising affinity membranes were extensively tested in dynamic experiments. The effects of operating parameters like feed concentration and flow rate on separation performances like binding capacity, selectivity and process yield have been studied in detail in order to find the optimal conditions for binding and elution steps. The membranes have been used over several complete chromatographic cycles to evaluate the effects of ageing and of membrane regeneration on dynamic binding capacity. A novel mathematical model is proposed that can describe all the chromatographic steps involved in the membrane affinity chromatography process for protein purification. The mathematical description is based on the species continuity equation coupled with a proper binding kinetic equation, and suitable to describe adequately the dispersion phenomena occurring both in the micro-porous membranes as well as in the extra-column devices used in the system. The model considers specifically all the different chromatographic steps, namely adsorption, washing and elution. The few relevant fitting parameters of the model were derived from a calibration with the experimental affinity cycles performed with pure IgG solutions, then the model is used to describe experimental data obtained in chromatographic cycles carried out with complex feeds as the cell culture supernatant. Simulations reveal a good agreement with experimental data in all the chromatography steps, both in the case of pure IgG solutions and for the cell culture supernatant considered.

Continuity and Change in New Products

This doctoral dissertation is triggered by an emergent problem: how can firms reinvent themselves? Continuity- and change-oriented decisions fundamentally shape overtime the activities and potential revenues of organizations and other adaptive systems, but both types of actions draw upon limited resources and rely on different organizational routines and capabilities. Most organizations appear to have difficulties in making tradeoffs, so that it is easier to overinvest in one of them than to successfully achieve a mixture of both. Nevertheless, theory and empirical evidence suggest that too little of either may reduce performance, indicating a need to learn more about how organizations reconcile these tensions. In the first paper, I moved from the consideration that rapid changes in competitive environments increasingly require firms to be “ambidextrous” implementing organizational mechanisms and structures that allow continuity- and change-oriented activities to be engaged at the same time. More specifically, I show that continuity- and change-related decisions can’t be confined either inside or outside the firm, but span overtime across distinct decision domains located within and beyond the organizational boundaries. Reconciling static and dynamic perspectives of ambidexterity, I conceptualize a firm’s strategy as a bundle of decisions about product attributes and components of the production team, proposing a multidimensional and dynamic model of structural ambidexterity that explains why and how firms could manage conflicting pressures for continuity and change in the context of new products. In the second study I note how rigorous systematic evidence documenting the success of ambidextrous organizations is lacking, and there has been very little investigation of how firms deal with continuity and change in new products. How to manage the transition form a successful product to another? What to change and what to keep? Incumbents that deal with series of products over time need to update their offerings in order to have the most relevant attributes to prospect clients without disappoint the current customer base. They need to both match and anticipate consumers’ preferences, blending something old with something new to satisfy the current demand and enlarge the herd by appealing to newer audiences. This paper contributes to strategic renewal and ambidexterity-related research with the first empirically assessment of a positive consumer response to ambidexterity in new products. Also, this study provides a practical method to monitor overtime the degree to which a brand or a firm is continuity- or change- oriented and evaluate different strategy profiles across two decision domains that play a pivotal role in new products: product attributes and components of the production team.

Essays in behavioral economics

In this work we conduct an experimental analysis on different behavioral models of economic choice. In particular, we analyze the role of overconfidence in shaping the beliefs of economics agents about the future path of their consumption or investment. We discuss the relevance of this bias in expectation formation both from a static and from a dynamic point of view and we analyze the effect of possible interventions aimed to achieve some policy goals. The methodology we follow is both theoretical and empirical. In particular, we make large use of controlled economic field experiments in order to test the predictions of the theoretical models we propose. In the second part of the thesis we discuss the role of cognition and personality in affecting economic preferences and choices. In this way we make a bridge between established psychological research and novel findings in economics. Finally, we conduct a field study on the role of incentives on education. We design different incentive schemes and we test, on randomized groups of students, their effectiveness in improving academic performance.

Rats' Learned Preferences for Flavors Encountered Early or Late in a Meal Paired with the Postingestive Effects of Glucose

Rats learn to prefer flavors that are followed by postingestive effects of nutrients. This experiment investigated whether the timing of a flavor (specifically, in the first or second half of the meal) influences learning about that flavor. Stronger learning about earlier or later flavorswould indicate when the rewarding postingestive effects of nutrients are sensed. Ratswith intragastric (IG) catheters drank saccharin-sweetened, calorically-dilute solutions with distinct flavors added, accompanied by IG infusion of glucose (+sessions) or water (−sessions). In both types of sessions, an “Early” flavorwas provided for the first 8 min and a “Late” flavor for the last 8 min. Thus, rats were trained with Early(+) and Late(+) in high-caloriemeals, and Early(−) and Late(−) in low-calorie meals. Strength of the learned preference for Early(+) and Late(+) was then assessed in a series of two-bottle choice tests between Early(+) vs. Early(−), Late(+) vs. Late(−), Early(+) vs. Late(+), and Early(−) vs. Late(−). Rats preferred both Early(+) and Late(+) over the respective (−) flavors. But Early(+)was only preferredwhen rats were tested hungry. Late(+) was preferred when rats were tested hungry or recently satiated. This indicates qualitatively different associations learned about flavors at different points in themeal.While not supporting the idea that postingestive effects become most strongly associated with later-occurring (“dessert”) flavors, it does suggest a reason dessert flavors may remain attractive in the absence of hunger.

Social Policy Facing the Problems of Youth Employment

Rumiana Stoilova (Bulgaria). Social Policy Facing the Problems of Youth Employment. Ms. Stoilova is a researcher in the Institute of Sociology in Sofia and worked on this project from October 1996 to September 1998. This project involved collecting both statistical and empirical data on the state of youth employment in Bulgaria, which was then compared with similar data from other European countries. One significant aspect was the parallel investigation of employment and unemployment, which took as a premise the continuity of professional experience where unemployment is just a temporary condition caused by external and internal factors. These need to be studied and changed on a systematic basis so as to create a more favourable market situation and to improve individuals' resources for improving their market opportunities. A second important aspect of the project was an analysis of the various entities active on the labour market, including government and private institutions, associations of unemployed persons, of employers or of trade unions, all with their specific legal powers and interests, and of the problems in communication between these. The major trends in youth unemployment during the period studied include a high proportion of the registered unemployed who are not eligible for social assistance, a lengthening of the average period of unemployment, an increase in the percentage of people who are unemployed for the first time and an increasing percentage of these who are not eligible for assistance, particularly among newly registered young people. At the same time the percentage of those for who work has been found is rising and during the last three years an increasing number of the unemployed have started some independent economic activity. Regional differences are also considerable and in the case of the Haskovo region represent a danger of losing the youngest generation, with resulting negative demographic effects. One major weakness of the existing institutional structure is the large scale of the black labour market, with clear negative implications for the young people drawn into it. The role of non-governmental organisations in providing support and information for the unemployed is growing and the government has recently introduced special preferences for organisations offering jobs to unemployed persons. Social policy in the labour market has however been largely restricted to passive measures, mostly because of the risk that poverty poses to people continuously excluded from the labour market. Among the active measures taken, well over half are concerned with providing jobs for the unemployed and there are very limited programmes for providing or improving qualifications. The nature of youth employment in Bulgaria can be seen in the influence of sustained structures (generation) and institutions (family and school). Ms. Stoilova studied the situation of the modern generation through a series of profiles, mostly those of continuously unemployed and self-employed persons, but also distinguishing between students and the unemployed, and between high school and university students. The different categories of young people were studied in separate mini-studies and the survey was carried out in five town in order to gather objective and subjective information on the state of the labour market in the different regions. She conducted interviews with several hundred young people covering questions of family background, career plans, attitudes to the labour situation and government measures to deal with it, and such questions as independence, mobility, attitude to work, etc. The interviews with young people unemployed for a long period of time show the risk involved in starting work and its link with dynamics of economic development. Their approval of structural reforms, of the financial restrictions connected with the introduction of a currency board and the inevitability of unemployment was largely declarative. The findings indicate that the continuously unemployed need practical knowledge and skills to "translate" the macroeconomic realities in concrete alternatives of individual work and initiative. The unemployed experience their exclusion from the labour market not only as a professional problem but also as an existential threat, of poverty, forced mobility and dependence on their parents' generation. The exclusion from the market of goods and services means more than just exercising restraint in their consumption, as it places restrictions on their personal development. Ms. Stoilova suggests that more efficient ways of providing financial aid and mobilisation are needed to counteract the social disintegration and marginalisation of the continuously unemployed. In measuring the speed of reform, university students took both employment opportunities and the implementation of the meritocratic principle in employment into account. When offered a hypothetical choice between a well-paid job and work in one's own profession, 62% would prefer opt for the well-paid job and for working for a company that offered career opportunities rather than employment in a family or own company. While most see the information gained during their studies as useful and interesting, relatively few see their education as competitive on a wider level and many were pessimistic about employment opportunities based on their qualifications. Very similar attitudes were found among high school students, with differences being due rather to family and personal situations. The unemployed, on the other hand, placed greater emphasis on possibilities of gaining or improving qualifications on a job and for the opportunities it would offer for personal contacts. High school students tend to attribute more significance to opportunities for personal accomplishment. A significant difference that five times fewer high school students were willing to work for state-owned companies, and many fewer expected to find permanent employment or to find a job in the area where they lived, Within the family situation, actual support for children seems to be higher than the feelings of confidence expressed in interviews. The attitudes of the families towards past experience seems to be linked with their ability to cope with the difficulties of the present, with those families which show an optimistic and active attitude towards the future having a greater respect for parents experience and tolerance in communication between parents and children.

Estimation of Transition Probabilities in Animal Carcinogenicity Experiments

This paper discusses estimation of the tumor incidence rate, the death rate given tumor is present and the death rate given tumor is absent using a discrete multistage model. The model was originally proposed by Dewanji and Kalbfleisch (1986) and the maximum likelihood estimate of the tumor incidence rate was obtained using EM algorithm. In this paper, we use a reparametrization to simplify the estimation procedure. The resulting estimates are not always the same as the maximum likelihood estimates but are asymptotically equivalent. In addition, an explicit expression for asymptotic variance and bias of the proposed estimators is also derived. These results can be used to compare efficiency of different sacrifice schemes in carcinogenicity experiments.

Preference of dairy cows for ryegrass, white clover and red clover, and its effects on nutrient supply and milk quality

Two experiments were conducted with 30 dairy cows each, to study the preference for fresh (Experiment 1) and ensiled (Experiment 2) ryegrass, white and red clover. Both experiments consisted of three choice diets with white or red clover or both, offered with ryegrass, and two diets with ryegrass mixed with white or red clover (40% clover). Cows consumed diets with 37.7% fresh white and 45.9% red clover, and no preference was observed when the cows were offered all three forages. By contrast, cows preferred white and red clover silage (73.0 and 69.2%, respectively) over ryegrass silage (of lower nutritive quality). When offered three forages, cows preferred white (59.8%) over red clover (17.5%) and ryegrass (22.7%). Choice diets resulted in diets similar (fresh forages) or higher in nutrient content and digestibility (silages). Treatments did not affect feed intake and performance. Choices compared to mixed diets with red clover silage were preferable regarding the fatty acid composition of the milk fat. Obviously, only large differences in nutrient and energy concentration facilitate preferences for clovers over ryegrass, which could, depending on clover type, be beneficial in terms of the milk's fatty acid composition.

Mapping Individual Variation in Male Mating Preference Space: Multiple Choice in a Color Polymorphic Cichlid Fish

Sexual selection theory largely rests on the assumption that populations contain individual variation in mating preferences and that individuals are consistent in their preferences. However, there are few empirical studies of within-population variation and even fewer have examined individual male mating preferences. Here, we studied a color polymorphic population of the Lake Victoria cichlid fish Neochromis omnicaeruleus, a species in which color morphs are associated with different sex-determining factors. Wild-caught males were tested in three-way choice trials with multiple combinations of different females belonging to the three color morphs. Compositional log-ratio techniques were applied to analyze individual male mating preferences. Large individual variation in consistency, strength, and direction of male mating preferences for female color morphs was found and hierarchical clustering of the compositional data revealed the presence of four distinct preference groups corresponding to the three color morphs in addition to a no-preference class. Consistency of individual male mating preferences was higher in males with strongest preferences. We discuss the implications of these findings for our understanding of the mechanisms underlying polymorphism in mating preferences.

Female preference for male color is necessary and sufficient for assortative mating in 2 cichlid sister species

A critical step for speciation in the face of gene flow is the origination of reproductive isolation. The evolution of assortative mating greatly facilitates this process. Assortative mating can be mediated by one or multiple cues across an array of sensory modalities. We here explore possible cues that may underlie female mate choice in a sympatric species pair of cichlid fish from Lake Victoria, Pundamilia pundamilia and Pundamilia nyererei. Previous studies identified species-specific female preferences for male coloration, but effects of other cues could not be ruled out. Therefore, we assessed female choice in a series of experiments in which we manipulated visual (color) and chemical cues. We show that the visibility of differences in nuptial hue (i.e., either blue or red) between males of the 2 species is necessary and sufficient for assortative mating by female mate choice. Such assortment mediated by a single cue may evolve relatively quickly, but could make reproductive isolation vulnerable to environmental changes. These findings confirm the important role of female mate choice for male nuptial hue in promoting the explosive speciation of African haplochromine cichlids.

Intraspecific sexual selection on a speciation trait, male coloration, in the Lake Victoria cichlid Pundamilia nyererei

The haplochromine cichlids of Lake Victoria constitute a classical example of explosive speciation. Extensive intra– and interspecific variation in male nuptial coloration and female mating preferences, in the absence of postzygotic isolation between species, has inspired the hypothesis that sexual selection has been a driving force in the origin of this species flock. This hypothesis rests on the premise that the phenotypic traits that underlie behavioural reproductive isolation between sister species diverged under sexual selection within a species. We test this premise in a Lake Victoria cichlid, by using laboratory experiments and field observations. We report that a male colour trait, which has previously been shown to be important for behavioural reproductive isolation between this species and a close relative, is under directional sexual selection by female mate choice within this species. This is consistent with the hypothesis that female choice has driven the divergence in male coloration between the two species. We also find that male territoriality is vital for male reproductive success and that multiple mating by females is common.

Interactions between dorsolateral and ventromedial prefrontal cortex underlie context-dependent stimulus valuation in goal-directed choice.

External circumstances and internal bodily states often change and require organisms to flexibly adapt valuation processes to select the optimal action in a given context. Here, we investigate the neurobiology of context-dependent valuation in 22 human subjects using functional magnetic resonance imaging. Subjects made binary choices between visual stimuli with three attributes (shape, color, and pattern) that were associated with monetary values. Context changes required subjects to deviate from the default shape valuation and to integrate a second attribute in order to comply with the goal to maximize rewards. Critically, this binary choice task did not involve any conflict between opposing monetary, temporal, or social preferences. We tested the hypothesis that interactions between regions of dorsolateral and ventromedial prefrontal cortex (dlPFC; vmPFC) implicated in self-control choices would also underlie the more general function of context-dependent valuation. Consistent with this idea, we found that the degree to which stimulus attributes were reflected in vmPFC activity varied as a function of context. In addition, activity in dlPFC increased when context changes required a reweighting of stimulus attribute values. Moreover, the strength of the functional connectivity between dlPFC and vmPFC was associated with the degree of context-specific attribute valuation in vmPFC at the time of choice. Our findings suggest that functional interactions between dlPFC and vmPFC are a key aspect of context-dependent valuation and that the role of this network during choices that require self-control to adjudicate between competing outcome preferences is a specific application of this more general neural mechanism.

Neural correlates of anticipation risk reflect risk preferences

Individual risk preferences have a large influence on decisions, such as financial investments, career and health choices, or gambling. Decision making under risk has been studied both behaviorally and on a neural level. It remains unclear, however, how risk attitudes are encoded and integrated with choice. Here, we investigate how risk preferences are reflected in neural regions known to process risk. We collected functional magnetic resonance images of 56 human subjects during a gambling task (Preuschoff et al., 2006). Subjects were grouped into risk averters and risk seekers according to the risk preferences they revealed in a separate lottery task. We found that during the anticipation of high-risk gambles, risk averters show stronger responses in ventral striatum and anterior insula compared to risk seekers. In addition, risk prediction error signals in anterior insula, inferior frontal gyrus, and anterior cingulate indicate that risk averters do not dissociate properly between gambles that are more or less risky than expected. We suggest this may result in a general overestimation of prospective risk and lead to risk avoidance behavior. This is the first study to show that behavioral risk preferences are reflected in the passive evaluation of risky situations. The results have implications on public policies in the financial and health domain.

Does the Use of Mobile Devices (Tablets and Smartphones) Affect Stated Choice Behavior in Web Surveys? An Application to Environmental Valuation

Web surveys are becoming increasingly popular in survey research. Compared with face-to-face, telephone and mail surveys, web surveys may contain a different and new source of measurement error and bias: the type of device that respondents use to answer the survey questions. To the best of our knowledge, this is the first study that tests whether the use of mobile devices affects survey characteristics and stated preferences in a web-based choice experiment. The web survey was carried out in Germany with 3,400 respondents, of which 12 per cent used a mobile device (i.e. tablet or smartphone), and comprised a stated choice experiment on externalities of renewable energy production using wind, solar and biomass. Our main finding is that survey characteristics such as interview length and acquiescence tendency are affected by the device used. In contrast to what might be expected, we find that, compared with respondents using desktop computers and laptops, mobile device users spent more time to answer the survey and are less likely to be prone to acquiescence bias. In the choice experiment, mobile device users tended to be more consistent in their stated choices, and there are differences in willingness to pay between both subsamples.