Mirko Perkusich

A model to detect problems on scrum-based software development projects

There is a high rate of software development projects that fails. Whenever problems can be detected ahead of time, software development projects may have better chances of success, and therefore save money and time. In this paper, we present a probabilistic model to help ScrumMasters to apply Scrum in organizations. The models goal is to provide information to the projects ScrumMaster to help him to be aware of the projects problems and have enough information to guide the team and improve the projects chances of success. We published a survey to collect data for this study and validated the model by applying it to scenarios. The results obtained so far show that the model is promising.

Early diagnosis of gastrointestinal cancer by using case-based and rule-based reasoning

We present a medical diagnosis decision support model for gastrointestinal cancer.The model uses case-based reasoning (CBR) and rule-based reasoning (RBR).We used real patient data as inputs to our model.The model was validated through K-fold cross validation and the paired t-test.Improved diagnosis accuracy compared to a CBR approach not using RBR in retrieval. In this paper, we present a medical diagnosis decision support model for gastrointestinal cancer. It should be used by general practitioners whenever there is a suspicion that a patient has this type of cancer. To build our model, we used Case-Based Reasoning (CBR) and Rule-Based Reasoning (RBR). We used real patient data as inputs to our model. We applied RBR to improve the CBR retrieve process. The models output presents the probability of the patient having a specific cancer. In order to adjust the attributes weights, we collected data from a general practitioner. To validate our model, we used K-fold cross validation and the paired t-test. The results showed that, with our approach, the accuracy of the diagnosis increased by 22.92% when compared to a CBR approach not using RBR in case retrieval. Furthermore, we evaluated our approach with an online questionnaire and semi-structured interviews. Even though, given the number of respondents, we cannot generalize our conclusions, the results indicate that our approach would be useful for general practitioners.

A Model-Based Approach to Support Validation of Medical Cyber-Physical Systems

Medical Cyber-Physical Systems (MCPS) are context-aware, life-critical systems with patient safety as the main concern, demanding rigorous processes for validation to guarantee user requirement compliance and specification-oriented correctness. In this article, we propose a model-based approach for early validation of MCPS, focusing on promoting reusability and productivity. It enables system developers to build MCPS formal models based on a library of patient and medical device models, and simulate the MCPS to identify undesirable behaviors at design time. Our approach has been applied to three different clinical scenarios to evaluate its reusability potential for different contexts. We have also validated our approach through an empirical evaluation with developers to assess productivity and reusability. Finally, our models have been formally verified considering functional and safety requirements and model coverage.

A model-based architecture for testing medical cyber-physical systems

Understanding the human body dynamics in response to any medical treatment makes automated decision support systems for healthcare quite complex. In this paper, we present an architecture for Medical Cyber-Physical Systems to help developers to generate test cases for their applications using models already validated. It is based on component models to simulate the operation of medical devices and patient data. Medical guidelines and a clinical database have been used together with statistical techniques to create regression models that simulate vital signs. A controlled experiment of a clinical scenario has been developed to validate the proposed architecture components. The results of this study indicate that models for the healthcare domain are a promising alternative to test their applications.

A Bayesian networks-based approach to assess and improve the teamwork quality of agile teams

Abstract CONTEXT: According to the agile principles and values, as well as recent research articles, teamwork factors are critical to achieve success in agile projects. However, teamwork does not automatically arise. There are some existing instruments with the purpose of assessing the teamwork quality based on Structural Equation Modeling (i.e., empirically derived) and Radar Plots, but they may not be useful in a concrete situation because these techniques are not advised for prediction and diagnosis purposes. OBJECTIVE: Analytically derive a Bayesian network model based on a literature review and a practitioner’s knowledge; and to assess its practical utility through a case study. METHOD: To build the model, we executed a top-down approach using data collected through a literature review and a domain practitioner. We assessed the model with a case study executed in three Scrum teams. RESULTS: Given the context of the case study, the model assists agile teams on assessing teamwork quality and identifying improvement opportunities, is easy to learn, and the cost-benefit for using it with the proposed procedure is positive. CONCLUSION: We concluded that we achieved promising results with the presented solution. However, it needs more evaluation and validation to generalize the obtained results.

Assisting the continuous improvement of Scrum projects using metrics and Bayesian networks.

Scrum is a simple process to understand, but hard to adopt. Therefore, there is a need for resources to assist on its adoption. In this paper, we present the process followed to build a Bayesian network to assist on the assessment of the quality of the software process in the context of Scrum projects. The model provides data to help Scrum Masters lead the improvement of business value delivery of Scrum teams. The process is divided into 2 phases. In the first phase, we built the Bayesian network based on expert knowledge extracted from the literature and experts. We used a top‐down approach and reasoning to define the key metrics necessary to build the models and their relationships. In the second phase, we updated the Bayesian network based on limitations of the first version. We validated the Bayesian network inferences with 10 simulated scenarios. Comparing both versions, for all scenarios, we improved the accuracy of the inferences. Therefore, we concluded that the Bayesian networks adequately represent Scrum projects from the viewpoint of the Scrum Master. Finally, the model built is in conformance with agile methods tailoring and can be adapted to any Scrum team.

Improving the Applicability of Bayesian Networks through Production Rules

One of the key challenges in constructing a Bayesian network BN is defining the node probability tables (NPT). For large-scale BN, learning NPT through domain experts knowledge elicitation is unfeasible. Previous works proposed solutions to this problem using the concept of ranked nodes; however, they have limited modeling capabilities or rely on BN experts to apply them, reducing their applicability. In this paper, we present an expert system based on production rules to define NPTs with the purpose of enabling the definition of NPTs by experts with no ranked nodes-specific knowledge. To create the rules, we elicited data from an expert in ranked nodes. To validate our approach, we executed an experiment with a BN already published in the literature to verify if, with our approach, a practitioner can achieve the same or better configuration for the NPTs. We used the Brier score to assess the NPTs accuracy and evaluated the results with the Wilcoxon test. All the Wilcoxon tests executed rejected the null hypotheses that stated that the Brier scores for the original NPTs method were the same as the new NPTs. By using our solution, a practitioner can accurately define NPTs without understanding the concept of ranked nodes.

A Gait Analysis Approach to Track Parkinson's Disease Evolution Using Principal Component Analysis

A research work is reproducible when all research artifacts such as as text, data, figure and code are available for independent researchers reproduce the results. In this paper, we present a reproducible gait analysis to track Parkinsons Disease evolution by monitoring walking abnormalities. Weapplied Principal Component Analysis into gait data to detect users abnormalities that may indicate the progression of Parkinsons Disease. We validated our approach with a public database of foot sensor data, which includes vertical ground reaction force records of subjects with healthy gait and Parkinsons Disease patients. We used the euclidean distance asdata classifier. We reached a classification accuracy of 81.00% with leave-one-out cross-validation, which demonstrates the feasibility of our approach for tracking PDs symptoms based on user gait. All relevant data to reproduce our results are available in a public web page.

A Bayesian Network Model to Assess Agile Teams' Teamwork Quality

The Agile Manifesto states that agile projects should focus on individuals and interactions over processes and tools. Agile teams are self-directed and considered one of the most valuable assets of the project. This requires the team members to be collaborative and embrace the concept of whole-team responsibility and commitment. Factors that influence teamwork quality such as team orientation, team leadership, and coordination in addition to highly specialized skills and corresponding division of work are barriers for achieving team effectiveness. To assist on the assessment of agile teamwork quality, in this paper, we present a Bayesian networks-based model. The model considers agile teams principles and industry best practices. We validated the model with simulated scenarios. The results are promising and encourage its usage to assess agile teams teamwork quality to promote continuous improvement.

Using Bayesian Network to estimate the value of decisions within the context of Value-Based Software Engineering

The software industrys current decision-making relating to product/project management and development is largely done in a value neutral setting, in which cost is the primary driver for every decision taken. However, numerous studies have shown that the primary critical success factor that differentiates successful products/projects from failed ones lie in the value domain. Therefore, to remain competitive, innovative and to grow, companies must change from cost-based to value-based decisionmaking where the decisions taken are the best for that companys overall value creation. This paper details a case study where value-based decisions made by key stakeholders to select features for the next sprint of an Internet of Things (IoT) project, stored in a decisions database, were used to build and validate a value estimation model. This models goal was to estimate the overall value contribution that each feature being discussed during a decision-making meeting would bring to the company, if selected for implementation. The estimation technique employed was Bayesian Network, and validation results were quite positive.