General Information

Data analytics teams utilize a wide range of tools to perform data modeling processes. When necessary, they also take an active role in developing artificial intelligence models through coding. All research details and hypotheses are designed by the researcher, and the analysis process is carried out in line with the researcher’s specific requests. Data analytics teams act as an analytical assistant supporting the researcher.

With the rapid expansion of data analytics tools, the need for analytical processes to be conducted by experienced professionals has increased significantly. This enables researchers to navigate complex and demanding analytical workflows more efficiently and reliably.

The Wistats platform provides a user-friendly interface for statistical analysis and machine learning models. It operates through seamless integration with widely accepted Python-based statistical and machine learning libraries in the background. Without requiring extensive technical expertise, analytical outputs can be generated efficiently. These libraries are explicitly stated as tools within the manuscript content.

Statistical and Machine Learning Methods Stated in the Manuscript

One of the following formats may be selected and included in the statistical and machine learning methods section of the manuscript.

Data Preprocessing Assistant

Data Preprocessing – General Information

After data preprocessing workflows are completed, statistical analysis models and machine learning models can be generated. Data preprocessing plays a critical role in ensuring that analyses are performed in a reliable, consistent, and scientifically valid manner.

Data preprocessing is performed by importing xlsx-formatted data files into projects created within the Wistats application. During this process, data structure, variable definitions, and analytical suitability are carefully evaluated.

Data preprocessing consultancy services are provided based on an man-hour service model, depending on project scope and data complexity.

Data Preprocessing Processes

• The data content is reviewed, and erroneous or inconsistent data areas are corrected within the xlsx file.
• The cleaned and structured data are imported into the relevant Wistats project.
• Variables and their types (categorical, numerical, etc.) are defined according to analytical requirements.
• Variables requested for analysis are examined using descriptive statistics.
• Categorical variables are grouped in a manner suitable for statistical analysis and modeling.
• Numerical variables are adjusted and structured to meet analysis and modeling requirements.

Statistical Analysis Assistant

Statistical Analysis – General Information

Various statistical tests are used to identify relationships between variables within a dataset. Statistical analysis enables researchers to explore data structure, compare groups, and interpret observed associations in a scientifically valid manner.

Descriptive statistical analyses are performed to summarize and characterize the overall properties of the data. Comparative statistical analyses are then conducted to evaluate differences and relationships between defined variables.

Statistical analysis consultancy services are provided based on a man-hour service model, depending on the scope and complexity of the analyses.

Statistical Analysis Processes

• Data preprocessing workflows are completed prior to statistical analysis.
• Categorical and numerical variables planned for analysis are defined within the Wistats application.
• Descriptive statistical analyses are performed for the entire dataset.
• Comparative statistical analyses are designed by defining inputs and outputs, and the analyses are executed and recorded accordingly.

Machine Learning Assistant

Machine Learning Algorithm – General Information

Machine learning algorithms can be used to predict or classify target outputs based on defined input variables. These algorithms enable the modeling of complex patterns and relationships within the data.

During the development of machine learning algorithms, open-source Python libraries (such as scikit-learn and statsmodels) are utilized, and modeling workflows are executed through the Wistats application interface.

Statistical analyses serve as a guiding step for selecting the most appropriate machine learning model. Therefore, relationships between variables are examined through statistical analysis prior to model development.

When developing machine learning algorithms, models with the highest accuracy and reliability metrics are identified and selected for evaluation.

The development of machine learning algorithms follows the CRISP-DM (Cross Industry Standard Process for Data Mining) methodology, which consists of the following steps:

• Business Understanding
• Data Understanding
• Data Preparation
• Modeling
• Evaluation
• Deployment

Machine learning consultancy services are provided based on a man-hour service model, depending on project scope and data complexity.

Machine Learning Algorithm – Processes

• Data preprocessing workflows are completed.
• Statistical analysis workflows are completed.
• Input and output variables are defined.
• Machine learning model training is performed.
• Model reliability metrics such as AUC (area under the curve), accuracy, sensitivity, and specificity are evaluated.
• The machine learning algorithm model is documented and reported.

WAI Web Application Development

WAI Web Application – General Information

Machine learning algorithms developed by the researcher can be deployed through a dedicated web application owned by the researcher or the research team. This approach transforms the developed models into interactive decision support tools rather than static academic outputs.

The developed web application can be linked within the manuscript, allowing journal editors and readers to access, explore, and test the models directly via the provided link. This linkage significantly enhances the practical and scientific value of the manuscript.

Information regarding this web application linkage may be stated in the cover letter using the following example format. The placeholder XXX refers to the researcher’s own domain name:

WAI Web Application Development – Processes

• The researcher purchases a domain name via a domain registration service such as GoDaddy.
• The purchased domain is redirected to WisdomEra DNS servers.
• An annual subscription license for the WAI web application within the WisdomEra ecosystem is purchased and maintained by the researcher.
• Web content and algorithm integration for the WAI application are developed by the DiTaKo team.
• The developed machine learning model is connected to web pages within the WAI application, enabling users to obtain real-time outputs based on input values.
• The generated web application links can be referenced within the manuscript.

WAI Web Product Link: https://wisdomera.io/urun-detayi