Search

Wireless sensor networks (WSNs) face significant challenges related to dependability, integrity, and confidentiality due to their widespread deployment. Intrusion detection emerges as a pivotal active defense technology to fortify the security of WSNs. Given the unique characteristics of WSNs, achieving a balance between precise data transmission and the limited energy resources of sensors, as well as between effective detection and optimal utilization of network resources, becomes imperative. In this research, we propose a fog computing-based intrusion detection system (IDS) tailored for a smart power grid environment. The primary focus of this article is to elucidate the application of IDS within the context of a smart grid. We present a novel stacked model based on ensemble learning algorithms, adept at accurately delineating the interconnections among fog nodes susceptible to cyber-attacks. To validate the efficacy of our approach, we conduct a series of comparative experiments utilizing the KDD-Cup-99 dataset with cross-validation. Our findings reveal that the Fog-IDS and Stacking-based approaches significantly enhance the performance of the IDS. Through simulations, we substantiate that our proposed approach effectively analyzes fog nodes using the deployed IDS, enabling the prediction of energy requirements at refill stations. Achieving an impressive accuracy rate exceeding 99.837%, MCC rate of 99.73%, and F1-score rate of 99.827%, our study underscores the robustness and reliability of the Fog-IDS and Stacking-based methods in augmenting IDS performance in WSNs. [ABSTRACT FROM AUTHOR]

In recent years, the world has dramatically moved toward using the internet of things (IoT), and the IoT has become a hot research field. Among various aspects of IoT, real-time cyber-threat protection is one of the most crucial elements due to the increasing number of cyber-attacks. However, current IoT devices often offer minimal security features and are vulnerable to cyber-attacks. Therefore, it is crucial to develop tools to detect such attacks in real time. This paper presents a new and intelligent network intrusion detection system named APAE that is based on an asymmetric parallel auto-encoder and is able to detect various attacks in IoT networks. The encoder part of APAE has a lightweight architecture that contains two encoders in parallel, each one having three successive layers of convolutional filters. The first encoder is for extracting local features using standard convolutional layers and a positional attention module. The second encoder also extracts the long-range information using dilated convolutional layers and a channel attention module. The decoder part of APAE is different from its encoder and has eight successive transposed convolution layers. The proposed APAE approach has a lightweight and suitable architecture for real-time attack detection and provides very good generalization performance even after training using very limited training records. The efficacy of the APAE has been evaluated using three popular public datasets named UNSW-NB15, CICIDS2017, and KDDCup99, and the results showed the superiority of the proposed model over the state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Hashtag feminism is a recent phenomenon, however, increasingly recurrent in the network society, connected by the Internet. One of the examples of this type of activism began in 2019 when a young woman named Mariana Ferrer used her social media to report that she had suffered sexual violence. Since then, her case has become increasingly known and several hashtags have been created to refer to the case, such as #JusticePorMariFerrer. Engagement with the hashtag reached a large number of engagements in November 2020, when the acquittal of the accused was reported and the term 'wrongful rape' was used by the portal that shared more details of the process. The indignation at what happened was so great that it invaded the streets of Brazil, with protests and demonstrations. Thus, the objective of this study is to identify the structure and aspects of the representation of feminism by hashtag, the semantic contexts, and common sense knowledge that permeate this type of manifestation. For this analysis, the KDD methodology was used, extracting 198,009 tweets with the hashtag #JustiçaPorMariFerrer in November 2020. Among the findings, the categories that reveal the discourse of rejection in relation to the system and digital solidarity stand out. Through the categories, it is possible to perceive that there is a distrust of official institutions and that women create connective solidarity networks. [ABSTRACT FROM AUTHOR]

Outlier detection is an important problem occurring in a wide range of areas. Outliers are the outcome of fraudulent behaviour, mechanical faults, human error, or simply natural deviations. Many data mining applications perform outlier detection, often as a preliminary step in order to filter out outliers and build more representative models. In this paper, we propose an outlier detection method based on a clustering process. The aim behind the proposal outlined in this paper is to overcome the specificity of many existing outlier detection techniques that fail to take into account the inherent dispersion of domain objects. The outlier detection method is based on four criteria designed to represent how human beings (experts in each domain) visually identify outliers within a set of objects after analysing the clusters. This has an advantage over other clustering‐based outlier detection techniques that are founded on a purely numerical analysis of clusters. Our proposal has been evaluated, with satisfactory results, on data (particularly time series) from two different domains: stabilometry, a branch of medicine studying balance‐related functions in human beings and electroencephalography (EEG), a neurological exploration used to diagnose nervous system disorders. To validate the proposed method, we studied method outlier detection and efficiency in terms of runtime. The results of regression analyses confirm that our proposal is useful for detecting outlier data in different domains, with a false positive rate of less than 2% and a reliability greater than 99%. [ABSTRACT FROM AUTHOR]

We employ the support vector machine (SVM) classifier, over different types of kernels, to investigate whether observable variables of individuals and their household information are able to describe their consumption decision of film at theaters in Brazil. Using a very big dataset of 340,000 individuals living in metropolitan areas of a whole large developing economy, we performed a Knowledge Discovery in Databases to classify the film consumers, which results in 80% instances correctly classified. To reduce the degrees of freedom for SVM and to learn the more important determinants of film consumption, we apply the Linear Discriminant Analysis that allows us to identify the key determinants of this consumption. The main individual characteristics are age, education (that merges to be a student), income, and preferences for cultural goods. Regarding the main geographic characteristics, these are the timing of sample, population concentration, and supply of movie theaters. The results point to an ineffective policy for the sector at the time investigated. [ABSTRACT FROM AUTHOR]

• Methodology to represent dynamic knowledge and explore emerging communities from corpora. • Analysis of communities and centrality over time is conducted, along with a proposed method for generating high-impact relations. • A dynamic knowledge graph is generated over the last nine years using the KDD papers corpus. • Emerging publications in KDD favor the creation of novel relations instead of entities. • Cohesion of the generated communities over time is constant, and the average community size grew over time. Knowledge graphs represent relationships between entities. These graphs can take dynamic forms to trace changes along time through text models and further used by reasoning systems with the intent to answer queries. In this research we explore their applicability for extracting temporal patterns of knowledge in the form of communities. To this end, we propose a method for generating knowledge relationships over unconnected components of a knowledge graph, allowing for a targeted exploration of emerging contents in corpora. This analysis is applied to the corpora of the Conference on Knowledge Discovery and Data Mining (KDD) publications over the last decade. We find the key knowledge communities over time and rank the underlying concepts. Results show that the publication efforts increasingly focus on graph research and the creation of relationships instead of new concepts. The acquired results confirm the validity of the proposed knowledge discovery methodology for community-centered analysis of emerging changes in dynamic knowledge graphs. [ABSTRACT FROM AUTHOR]

Different classical and hybrid intelligent techniques have evolved over the years for the purpose of clustering different datasets. In the machine learning environment, among the different techniques in vogue, we interchangeably used the terms classification, regression and clustering for the inductive learning. The DIKW process of knowledge engineering has been extended and influenced by different intelligent techniques, hybrid in nature. In this article, we propose a hybrid model of Fuzzy C Means with K-nearest neighbour(FCM-KNN) to classify different datasets. In the proposed model, we used one input layer, two middle layers for processing the input files and one output layer for visualization and data collection. We compare the same with Naive Bayes Learning and SVM. Out of three discussed machine learning techniques, Naive Bayes and SVM are supervised in nature and FCM-KNN is an unsupervised classification techniques. Results of the proposed models are demonstrated on a real-life dataset using KNIME model design. A comparative study with the results obtained with the proposed architecture outperforms the other two models, viz. Naive Bayes and SVM, in terms of correct computation percentage. [ABSTRACT FROM AUTHOR]

The purpose of this research is to describe the process of knowledges discovery applied to the results of the surveys accomplished in the characterization process of a student brigade of the career Engineering in Informatics Sciences. For which was applied the KDD algorithm steps, obtaining 5 subgroups of students with similar characteristics and well defined among its members. Allowing get a clearer view of the objectives to define in the educational project, as well as make proper planning of activities based on the specific goals to be achieved by each subgroup; in order to ensure proper interrelationship between the different factors involved in the success and competitiveness of enterprises that employ graduates the University Informatics Sciences. [ABSTRACT FROM AUTHOR]

Clustering techniques find homogeneous and distinguishable prototypes. Careful interpretation of these prototypes is crucial to assist the experts to better organize this know-how and to really improve their decision-making processes. The Traffic Lights Panel was introduced in 2009 as a postprocessing tool to provide understanding of clustering prototypes. In this work, annotated Traffic Lights Panel (aTLP) is presented as an enrichment of the TLP to manage the intrinsic uncertainty related with prototypes themselves. The aTLP handles uncertainty through a quantification of the prototypes' purity based on the variation coefficients (VC) and an associated color-based uncertainty model, with two dimensions - tone and saturation - representing nominal trend and purity of the prototype. An application to a waste-water treatment plant in Slovenia, in a discrete and continuous approach, suggests that aTLP seems a useful and friendly tool able to reduce the gap between data mining and effective decision support, towards informed-decisions. [ABSTRACT FROM AUTHOR]

Abstract: Without data preparation, data mining algorithms cannot operate on data within the knowledge discovery in databases (KDD) process. In fact, the success of later KDD phases largely depends on the data preparation stage. The use of mechanisms for automatically preparing data saves a lot of time and resources within the KDD process. These resources will then be available for use at later, less automatable stages, for example, during results interpretation. We have proposed a general-purpose mechanism applicable to multiple domains in order to improve the data preparation phase in the KDD process. This mechanism processes and automatically converts input data to a suitable format for the application of different data preparation techniques based on a known syntax. It is based on the use of description logic. Taking a generic UML2 data model as a reference, this mechanism is able to check whether any XML data source whatsoever can be transformed and modelled as a subsumption or instance of the above UML2 model. Thus it automatically identifies a consistent, non-ambiguous and finite set of XLST transformations which are used to prepare the data for the application of data mining techniques, obviating the need to expend resources on the preliminary preparation and formatting stage. The proposed mechanism was applied on structurally complex data from four different domains. In order to test the validity of the proposal, we have applied data mining techniques to extract knowledge from the prepared data. The sound results of applying our proposal to several different domains confirm that it is applicable to any XML data source, as well as being correct, computationally efficient and saving time during the data preparation phase. [Copyright &y& Elsevier]

Abstract: Temporal Data Mining is a core concept of Knowledge Discovery in Databases handling time-oriented data. State-of-the-art methods are capable of preserving the temporal order of events as well as the temporal intervals in between. The temporal characteristics of the events themselves, however, can likely lead to numerous uninteresting patterns found by current approaches. We present a new definition of the temporal characteristics of events and enhance related work for pattern finding by utilizing temporal relations, like meets, starts, or during, instead of just intervals between events. These prerequisites result in MEMuRY, a new procedure for Temporal Data Mining that preserves and mines additional time-oriented information. Our procedure is supported by SAPPERLOT, an interactive visual interface for exploring the patterns. Furthermore, we illustrate the efficiency of our procedure presenting a benchmark of the procedure's run-time behavior. A usage scenario shows how the procedure can provide new insights. [Copyright &y& Elsevier]

Abstract: In recent years, much attention has been given to the problem of outlier detection, whose aim is to detect outliers – objects who behave in an unexpected way or have abnormal properties. The identification of outliers is important for many applications such as intrusion detection, credit card fraud, criminal activities in electronic commerce, medical diagnosis and anti-terrorism, etc. In this paper, we propose a hybrid approach to outlier detection, which combines the opinions from boundary-based and distance-based methods for outlier detection (). We give a novel definition of outliers –BD (boundary and distance)-based outliers, by virtue of the notion of boundary region in rough set theory and the definitions of distance-based outliers. An algorithm to find such outliers is also given. And the effectiveness of our method for outlier detection is demonstrated on two publicly available databases. [Copyright &y& Elsevier]

The purpose of this paper is to study the effectiveness of data imputation methods in dealing with data missingness in the data mining phase of knowledge discovery in Database (KDD). The application of data mining techniques without careful consideration of missing data can result into biased results and skewed conclusions. This research explores the impact of data missingness at various levels in KDD models employing neural networks as the primary data mining algorithm. Four of the most commonly utilized data imputation methods - Case Deletion, Mean Substitution, Regression Imputation, and Multiple Imputation were evalutated using Root Mean Square (RMS) Values, ANOVA Testing, T-tests, and Tukey's Honestly Significant Difference Test to assess the differences of performance levels between various Knowledge Discovery and Neural Network Models, both in the presence and absence of Missing Data. [ABSTRACT FROM AUTHOR]

Nowadays, one of the greatest challenges of every educational institution is to retain students without prejudice to academic quality. In universities, statistics show that the academic periods where there is greater dropout rate correspond to the first semesters (Tello, 2007). On this topic several studies have been developed that provide dropout statistics and on which possible causes, such as deficiencies in basic concepts from secondary education and methods of study, among others are suggested. In this sense, this work addressed this issue through an analysis of data from a university academic basis using the data mining technique more suitable for the study (of screening conducted among various techniques) which offers a variety of statistical and computational methods to investigate the existence of relationships and "hidden" behavior patterns in freshmen dtudents.. These relationships may suggest causal explanations that can be verified later or action strategies to achieve some objectives of change and support for decision making. For this case the design of a prototype software by means of a neural network using Kohonen self-organizational maps, and an artificial intelligence technique [ABSTRACT FROM AUTHOR]

Abstract: “One person’s noise is another person’s signal” (Knorr, E., Ng, R. (1998). Algorithms for mining distance-based outliers in large datasets. In Proceedings of the 24th VLDB conference, New York (pp. 392–403)). In recent years, much attention has been given to the problem of outlier detection, whose aim is to detect outliers – objects which behave in an unexpected way or have abnormal properties. Detecting such outliers is important for many applications such as criminal activities in electronic commerce, computer intrusion attacks, terrorist threats, agricultural pest infestations, etc. And outlier detection is critically important in the information-based society. In this paper, we discuss some issues about outlier detection in rough set theory which emerged about 20 years ago, and is nowadays a rapidly developing branch of artificial intelligence and soft computing. First, we propose a novel definition of outliers in information systems of rough set theory –sequence-based outliers. An algorithm to find such outliers in rough set theory is also given. The effectiveness of sequence-based method for outlier detection is demonstrated on two publicly available databases. Second, we introduce traditional distance-based outlier detection to rough set theory and discuss the definitions of distance metrics for distance-based outlier detection in rough set theory. [Copyright &y& Elsevier]

Proper management and implementation of an effective child welfare agency requires the constant use of information about the experiences and outcomes of children involved in the system, emphasizing the need for comprehensive, timely, and accurate data. In the past 20 years, there have been many advances in technology that can maximize the potential of administrative data to promote better evaluation and management in the field of child welfare. Specifically, this article discusses the use of knowledge discovery and data mining (KDD), which makes it possible to create longitudinal data files from administrative data sources, extract valuable knowledge, and make the information available via a user-friendly public Web site. This article demonstrates a successful project in North Carolina where knowledge discovery and data mining technology was used to develop a comprehensive set of child welfare outcomes available through a public Web site to facilitate information sharing of child welfare data to improve policy and practice. [ABSTRACT FROM AUTHOR]

“One person's noise is another person's signal” (Knorr and Ng 1998). In recent years, much attention has been given to the problem of outlier detection, whose aim is to detect outliers—objects who behave in an unexpected way or have abnormal properties. Detecting such outliers is important for many applications such as criminal activities in electronic commerce, computer intrusion attacks, terrorist threats, agricultural pest infestations. In this paper, we suggest to exploit the framework of rough sets for detecting outliers. We propose a novel definition of outliers—RMF (rough membership function)-based outliers, by virtue of the notion of rough membership function in rough set theory. An algorithm to find such outliers is also given. And the effectiveness of RMF-based method is demonstrated on two publicly available data sets. [ABSTRACT FROM AUTHOR]

Abstract: Case-based reasoning (CBR) is a paradigm, concept and instinctive mechanism for problem solving. Recently, CBR has been widely integrated with some AI algorithms and applied to various kinds of problems. The ill-defined and unstructured problems are suitably solved by CBR. This research proposes a hybrid CBR mechanism including two stages. In stage I, the genetic algorithm is adopted to improve efficiency of case retrieving process. Compared to traditional CBR, the proposed mechanism could reduce about 14% case evaluations, but still achieved 90% satisfactory results. In stage II, the knowledge discovering and data mining (KDD) processes are implemented to produce the refined information from the retrieved cases. Because these retrieved cases and target problem satisfy similar or even same conditions, the outcome of KDD would be more valuable for reference. In addition to retrieved cases of stage I, the proposed mechanism provides direction and relevant knowledge for decision makers in their decision supporting and revising processes in stage II. The proposed CBR mechanism also deals with efficiency and outcome quality issues of traditional CBR. [Copyright &y& Elsevier]

Knowledge discovery and data mining methods have been successful in many domains. However, their abilities to build or discover a domain theory remain unclear. This is largely due to the fact that many fundamental KDD&DM methodological questions are still unexplored such as (1) the nature of the information contained in input data relative to the domain theory, and (2) the nature of the knowledge that these methods discover. The goal of this paper is to clarify methodological questions of KDD&DM methods. This is done by using the concept of Relational Data Mining (RDM), representative measurement theory, an ontology of a subject domain, a many-sorted empirical system (algebraic structure in the first-order logic), and an ontology of a KDD&DM method. The paper concludes with a review of our RDM approach and 'Discovery' system built on this methodology that can analyze any hypotheses represented in the first-order logic and use any input by representing it in many-sorted empirical system. [ABSTRACT FROM AUTHOR]

Current Business Intelligence (BI) initiatives customize DM-KDD techniques into business analytics, which cannot be used in applications other than business. A review of current methodology at the strategic level of the KM/KDD domains indicates that there exists no general formal framework which can be adopted in new applications, or new application areas. There are no established procedures for the domain expert to express their prior knowledge, understanding and aims in a way which can be linked to KDD/DMM processes and subsequent deployment of discovered knowledge. It is suggested that the sequential life-cycle project-management approach of CRISP-DM needs to be complemented by a dynamic interactive view of a conceptual data/information/knowledge hierarchy in the KM context. It is also suggested that a graphical/visual knowledge representation framework needs to be developed as the basis of a knowledge and discovery and communication framework (KDCF). A review of the limitations in DM methodology at the technical/technological level leads to the conclusion that there is no coherent DM methodology to guide the choice of models and their evaluation, that the DM discipline is fractionated, and that the fundamental search and sampling paradigms have been insufficiently utilized in DM development. It is proposed that development of linked data and model ontologies, together with a DM-epistemology, and associated with full exploitation of search and sampling could lead to improved cohesion and efficacy of the DM discipline. [ABSTRACT FROM AUTHOR]

Currently statistical and artificial neural network methods dominate in data mining applications. Alternative relational (symbolic) data mining methods have shown their effectiveness in robotics, drug design, and other areas. Neural networks and decision tree methods have serious limitations in capturing relations that may have a variety of forms. Learning systems based on symbolic first-order logic (FOL) representations capture relations naturally. The learned regularities are understandable directly in domain terms that help to build a domain theory. This paper describes relational data mining methodology and develops it further for numeric data such as financial and spatial data. This includes (1) comparing the attribute-value representation with the relational representation, (2) defining a new concept of joint relational representations, (3) a process of their use, and the Discovery algorithm. This methodology handles uniformly the numerical and interval forecasting tasks as well as classification tasks. It is shown that Relational Data Mining (RDM) can handle multiple constrains, initial rules and background knowledge very naturally to reduce the search space in contrast with attribute-based data mining. Theoretical concepts are illustrated with examples from financial and image processing domains. [ABSTRACT FROM AUTHOR]

By using data mining techniques, the data stored in a Data Warehouse (DW) can be analyzed for the purpose of uncovering and predicting hidden patterns within the data. So far, different approaches have been proposed to accomplish the conceptual design of DWs by following the multidimensional (MD) modeling paradigm. In previous work, we have proposed a UML profile for DWs enabling the specification of main MD properties at conceptual level. This paper presents a novel approach to integrating data mining models into multidimensional models in order to accomplish the conceptual design of DWs with Association Rules (AR). To this goal, we extend our previous work by providing another UML profile that allows us to specify Association Rules mining models for DW at conceptual level in a clear and expressive way. The main advantage of our proposal is that the Association Rules rely on the goals and user requirements of the Data Warehouse, instead of the traditional method of specifying Association Rules by considering only the final database implementation structures such as tables, rows or columns. In this way, ARs are specified in the early stages of a DW project, thus reducing the development time and cost. Finally, in order to show the benefits of our approach, we have implemented the specified Association Rules on a commercial database management server. [Copyright &y& Elsevier]

Classification Association Rule Mining (CARM) systems operate by applying an Association Rule Mining (ARM) method to obtain classification rules from a training set of previously classified data. The rules thus generated will be influenced by the choice of ARM parameters employed by the algorithm (typically support and confidence threshold values). In this paper we examine the effect that this choice has on the predictive accuracy of CARM methods. We show that the accuracy can almost always be improved by a suitable choice of parameters, and describe a hill-climbing method for finding the best parameter settings. We also demonstrate that the proposed hill-climbing method is most effective when coupled with a fast CARM algorithm such as the TFPC algorithm which is also described. [Copyright &y& Elsevier]

Abstract: This paper tackles problems encountered in mining of incomplete data for knowledge discovery of construction databases. As historical construction data are expensive and time-consuming to collect, any waste of incomplete data means not only loss of knowledge but also increase of costs for knowledge discovery of construction engineering. Unfortunately, incompleteness is omnipresent in the existing construction databases. This paper proposes a VaFALCON (Variable-Attribute Fuzzy Adaptive Logic Control Network) neuro-fuzzy system that is based on the architecture of the original FALCON and equipped with capabilities for mining incomplete data. Three real world examples are selected to test the proposed VaFALCON. The testing results show that the proposed VaFALCON is able to improve the system accuracy up to 84.5% and recover accuracy at least 81% even under the severe data incompleteness case, where all datasets of the database are incomplete. [Copyright &y& Elsevier]

A successful application of data mining to bioinformatics is protein classification. A number of techniques have been developed to classify proteins according to important features in their sequences, secondary structures, or three-dimensional structures. In this paper, we introduce a novel approach to protein classification based on significant patterns discovered on the surface of a protein. We define a notion called α-surface. We discuss the geometric properties of a-surface and present an algorithm that calculates the a-surface from a finite set of points in W. We apply the algorithm to extracting the a-surface of a protein and use a pattern discovery algorithm to discover frequently occurring patterns on the surfaces. The pattern discovery algorithm utilizes a new index structure called the ΔB+ tree. We use these patterns to classify the proteins. While most existing techniques focus on the binary classification problem, we apply our approach to classifying three families of proteins. Experimental results show the good performance of the proposed approach. [ABSTRACT FROM AUTHOR]

Nowadays, large amounts of data are generated in the medical domain. Various physiological signals generated from different organs can be recorded to extract interesting information about patients' health. The analysis of physiological signals is a hard task that requires the use of specific approaches such as the Knowledge Discovery in Databases process. The application of such process in the domain of medicine has a series of implications and difficulties, especially regarding the application of data mining techniques to data, mainly time series, gathered from medical examinations of patients. The goal of this paper is to describe the lessons learned and the experience gathered by the authors applying data mining techniques to real medical patient data including time series. In this research, we carried out an exhaustive case study working on data from two medical fields: stabilometry (15 professional basketball players, 18 elite ice skaters) and electroencephalography (100 healthy patients, 100 epileptic patients). We applied a previously proposed knowledge discovery framework for classification purpose obtaining good results in terms of classification accuracy (greater than 99% in both fields). The good results obtained in our research are the groundwork for the lessons learned and recommendations made in this position paper that intends to be a guide for experts who have to face similar medical data mining projects. [ABSTRACT FROM AUTHOR]

• We assess benefits of MODIS LST for predicting corn yield across the US Corn Belt. • LST can predict annual corn yield with less error than air temperature. • Improvements with LST remained when adjusting for common meteorological factors. • LST outperformed air temperature by a much wider margin in 2012, a drought year. While canopy temperature has been extensively utilized for field-level crop health assessment, the application of satellite-based land surface temperature (LST) images for corn yield modeling has been limited. Furthermore, long term yield projections in the context of climate change have primarily employed air temperature (Tair) and precipitation, which may inadequately reflect crop stress. This study assessed potential benefits of satellite-derived LST for predicting annual corn yield across the US Corn Belt from 2010 to 2016. A novel killing degree day metric (LST KDD) was computed with daily LST images from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor and compared to the typically used Tair-based metric (Tair KDD). Our findings provide strong evidence that LST KDD is capable of predicting annual corn yield with less error than Tair KDD (R2/RMSE of 0.65/15.3 Bu/Acre vs. 0.56/17.2 Bu/Acre). Even while adjusting for seasonal temperature and precipitation parameters, the R2 and RMSE of the LST model were approximately 9% higher and 2.0 Bu/Acre lower than the Tair model, respectively. The superior performance of LST can be attributed to its ability to better incorporate evaporative cooling and water stress. We conclude that MODIS LST can improve yield forecasts several months prior to harvest, especially during extremely warm and dry growing seasons. Furthermore, the better performance of LST models over Tair and precipitation models suggest that subsequent long term yield projections should consider additional factors indicative of water stress. [ABSTRACT FROM AUTHOR]