[pt] Problemas de classificação foram amplamente estudados na literatura de aprendizado de máquina, gerando aplicações em diversas áreas. No entanto, em diversos cenários, custos por erro de classificação podem variar bastante, o que motiva o estudo de técnicas de classificação sensível ao custo. Nesse trabalho, discutimos o uso de árvores de decisão para o problema mais geral de Aprendizado Sensível ao Custo do Exemplo (ASCE), onde os custos dos erros de classificação variam com o exemplo. Uma das grandes vantagens das árvores de decisão é que são fáceis de interpretar, o que é uma propriedade altamente desejável em diversas aplicações. Propomos um novo método de seleção de atributos para construir árvores de decisão para o problema ASCE e discutimos como este pode ser implementado de forma eficiente. Por fim, comparamos o nosso método com dois outros algoritmos de árvore de decisão propostos recentemente na literatura, em 3 bases de dados públicas. / [en] Classification problems have been widely studied in the machine learning literature, generating applications in several areas. However, in a number of scenarios, misclassification costs can vary substantially, which motivates the study of Cost-Sensitive Learning techniques. In the present work, we discuss the use of decision trees on the more general Example-Dependent Cost-Sensitive Problem (EDCSP), where misclassification costs vary with each example. One of the main advantages of decision trees is that they are easy to interpret, which is a highly desirable property in a number of applications. We propose a new attribute selection method for constructing decision trees for the EDCSP and discuss how it can be efficiently implemented. Finally, we compare our new method with two other decision tree algorithms recently proposed in the literature, in 3 publicly available datasets.

Lung cancer is one the most dangerous types of all cancer. Several studies have explored the use of machine learning methods to predict and diagnose this cancer. This study explored the potential of decision tree (DT) and random forest (RF) classification models, in the context of a small transcriptome dataset for outcome prediction of different subtypes on lung cancer. In the study we compared the three subtypes; adenocarcinomas (AC), small cell lung cancer (SCLC) and squamous cell carcinomas (SCC) with normal lung tissue by applying the two machine learning methods from caret R package. The DT and RF model and their validation showed different results for each subtype of the lung cancer data. The DT found more features and validated them with better metrics. Analysis of the biological relevance was focused on the identified features for each of the subtypes AC, SCLC and SCC. The DT presented a detailed insight into the biological data which was essential by classifying it as a biomarker. The identified features from this research may serve as potential candidate genes which could be explored further to confirm their role in corresponding lung cancer types and contribute to targeted diagnostics of different subtypes.

Hot Mix Asphalt (HMA) overlay is a major rehabilitation treatment for the existing deteriorated pavements (both flexible and rigid pavements). Reflective cracking (RC) is the most common distress type appearing in the HMA overlays which structurally and functionally degrades the whole pavement structure, especially under high traffic volume. Although many studies have been conducted to identify the best performing RC mitigation technique, the level of success varies from premature failure to good performance in the field. In Florida, Asphalt Rubber Membrane Interlayers (ARMIs) have been used as a RC mitigation technique but its field performance has not been successful. In this study, the best performing means to mitigate RC in the overlays considering Florida’s special conditions have been investigated. The research methodology includes (1) extensive literature reviews regarding the RC mechanism and introduced mitigation options, (2) nationwide survey for understanding the current practice of RC management in the U.S., and (3) the development of decision trees for the selection of the best performing RC mitigation method. Extensive literature reviews have been conducted to identify current available RC mitigation techniques and the advantages and disadvantages of each technique were compared. Lesson learned from the collected case studies were used as input for the selection of the best performing RC mitigation techniques for Florida’s roads. The key input parameters in selecting optimum mitigation techniques are: 1) overlay characterization, 2) existing pavement condition, 3) base and subgrade structural condition, 4) environmental condition and 5) traffic level. In addition, to understand the current iv practices how reflective cracking is managed in each state, a nationwide survey was conducted by distributing the survey questionnaire (with the emphasis on flexible pavement) to all other highway agencies. Based on the responses, the most successful method of treatment is to increase the thickness of HMA overlay. Crack arresting layer is considered to be in the second place among its users. Lack of cost analysis and low rate of successful practices raise the necessity of conducting more research on this subject. Considering Florida’s special conditions (climate, materials, distress type, and geological conditions) and the RC mechanism, two RC mitigation techniques have been proposed: 1) overlay reinforcement (i.e. geosynthetic reinforcement) for the existing flexible pavements and 2) Stress Absorbing Membrane Interlayer (SAMI) for the existing rigid pavements. As the final products of this study, decision trees to select an optimum RC mitigation technique for both flexible and rigid pavements were developed. The decision trees can provide a detailed guideline to pavement engineer how to consider the affecting parameters in the selection of RC mitigation technique.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.