Strategies and Methods for Analyzing Multivariate-Multidimensional Data Resulted from Agricultural Research

Dear Colleagues,

Traditionally, data analysis in agricultural research has been “dominated” by univariate statistical approaches such as analysis of variance (ANOVA). ANOVA has facilitated the analysis of experimental data in a coherent manner, and this was largely acceptable. Today, however, the scope of agricultural research and practice produces massive and “sophisticated” datasets that are multivariate and multidimensional in nature. Such complicated and large datasets necessitate “sophisticated” analyses in conjunction with traditional analyses such as ANOVA. As agricultural systems become more complicated and as new technology in data collection advances, it has become a necessity for the researchers to use more multivariate and multidimensional statistical and data analysis approaches capable of exploiting the complexity of the data. Spatial analysis is also becoming indispensable, as location-based datasets often carry critical information that requires appropriate handling to ensure meaningful insights, or they offer additional information that is usually omitted when data are treated as a list of values and not as measurements with specific positions.

Background and history

In recent years, agricultural sciences have witnessed global changes in the collection, processing and analysis of data. Agricultural experimentation, in its beginning stages, only needed ANOVA and other similar univariate methods to satisfy more or less simple and few variables integrating relatively simple research questions. However, with technological advancements such as phenotyping, precision agriculture and environmental monitoring, data volume has increased dramatically and their complexity has risen. Few and specific interactions focusing on a particular set of data are not sufficient with today’s large datasets, which capture a wider range of complex interactions and relationships.

To fill this gap, agronomists have been using more Factor Analysis, Principal Components Analysis, Cluster Analysis, Canonical Correlation Analysis, Multidimensional Scaling, Correspondence Analysis, Structural Equation Models and even advanced machine learning algorithms, which are multidimensional and multivariate in nature. In addition, spatial analysis that integrates spatial coordinates into these analyses allows researchers to identify geographic patterns, supporting decision-making that accounts for spatial variability.

These techniques enable researchers to explore deeper into the relationships among a variety of variables and data types (e.g., nominal, ordinal, scale and circular) that are pertinent in any agricultural system.

Aim and Scope

Innovation-related methods that extend traditional ANOVA methods are demonstrated in this Special Issue in order to explore new horizons from the perspective of agricultural data. We hope for the collection to be beneficial for agricultural researchers who are looking for strategies to analyze complex, high-dimensional data. This Special Issue offers mappings on diverse topics, including classical multivariate statistical methods and advanced computational ones. The purpose is to assist the agricultural research community in coming up with more effective strategies to manage the large amounts of data that have quickly increased due to their research activities. This Special Issue aims to enhance univariate analyses so that the more dynamic nature of agricultural systems can be better explained and understood. Additionally, exploring methodologies such as Causal Inference, which leverages spatial data alongside multidimensional variables, highlights the importance of spatially informed analyses in understanding causal relationships and improving agricultural practices. It is expected that the solution of problems relative to data reduction and predictive modeling of multivariate relationships or integrative approaches will have spatial, time and hierarchical dimensions.

Cutting-edge Research

Agricultural research today can be characterized as a crossing of boundaries that combines genetics with environmentalism, remote sensing and several other areas. These data are complicated and therefore need, in addition or complementary to traditional statistical methods, contemporary methods or strategies to manage or analyze them. Recent developments in multivariate approaches and machine learning algorithms, including Random Forests, Support Vector Machines, Partial Least Squares Regression and Structural Equation Modeling (SEM), have also created opportunities for the analysis of agricultural data. Spatial modeling and analysis are key components in this regard, providing insights into spatially dependent patterns that can enhance predictive accuracy and resource allocation. Further contributions include the advancements in artificial intelligence and machine learning that have given effective tools and techniques for recognizing patterns, for prediction, for classification and for interpreting data. With these techniques, it is possible to model very complicated relationships and increase the degree of predictive accuracy. The contributions presented in this Special Issue will concentrate on using these “sophisticated” techniques to tackle practical questions in agriculture such as increasing agricultural output (products), natural resource management and sustainable development in a changing climate.

Conclusion

To cope with today’s agricultural research and its complexities, a transition into more advanced multivariate statistical techniques is necessary. Within this Special Issue, attention is paid to the changing research landscape through the lens of new methods and strategies of analyzing the available data. These suggestions and practices are based on much more varied strategies, which are outside but complementary to the conventional measure of ANOVA, in order to have a comprehensive picture of agricultural systems and improve decision-making. We also encourage contributions that integrate spatial analysis into multivariate frameworks to address real-world challenges. We welcome submissions demonstrating advanced methods, cross-disciplinary strategies and innovative practices that are at the forefront of the development of data analysis processes in agriculture. Papers using new strategies for traditional statistical analyses are also appreciated.

Dr. George Menexes
Dr. Thomas Koutsos
Dr. Angelos Markos
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Agronomy is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• multivariate and multidimensional data analysis
• big data
• design of experiments