Traditional Artificial Intelligence Research Articles

Research on adaptive conversion of AI language based on rough set

In order to solve the problems of high complexity and low computational efficiency in traditional artificial intelligence (AI) language conversion methods, an adaptive AI language conversion method based on rough set is proposed. AI language preprocessing is realised by pre-emphasising, adding window, frame processing and endpoint detection. The attribute reduction algorithm based on rough set theory is used to select the features of AI language. The dimension of input feature vector is reduced. The experimental results show that after feature extraction, the computational efficiency is obviously improved, and the efficiency of the proposed method is the highest, averaging close to 100%. Compared with the traditional method, the complexity of the proposed method is lower, and the average complexity is 1.68% during the ten experimental iterations. This method simplifies the adaptive conversion process of AI language and has high computational efficiency.

Tool failure recognition using inconsistent data

Data is everything - at least this is one of the main messages of the ongoing industrial revolution. Manufacturing companies all over the world are expanding their digital infrastructure and knowledge on data analysis in the hope of increasing their KPIs with the help of artificial intelligence (AI). Although several well-designed data-driven solutions are available, the most crucial part, data preparation is still not fully supported. In this paper a framework is presented for processing sensor data of machining processes with variable cycle times in an unstable environment. Traditional and novel AI algorithms are tested on the data of a vulcanization process from the automotive industry, namely from tire manufacturing’s curing phase. The process in question consists of several subprocesses, and the quality of curing is mostly dependent of the status of a specific type of machine tool. Conventional methods (e.g., examining the cured product manually) are currently used for failure recognition, however the examination is only feasible after a long delay due to the extreme level of heat, which leads to unnecessary and unwanted scrap production. Therefore, a more sophisticated and complex approach is required to increase quality score. A combination of mathematical methods is proposed combining t-SNE feature representation, convolutional neural network, and linear programming optimization. The model highly relies on the tool’s continuous degradation characteristics. The threshold for the given binary classification is set by maximizing the accuracy of the detection model. The main contribution of the research is the method of inconsistent sensor data manipulation which supports a unique combination of AI models for early failure recognition.

Sensor network prediction based on spatial and temporal GNN

Multi-sensor prediction is a hotspot for research and development in sensor management technologies. Thanks to artificial intelligence, researchers have been able to effectively use neural networks and traditional artificial intelligence approaches to multi-sensor prediction in recent years. In this model, we try to present the sensors network as an unweighted graph, based on the GNN with spatial and temporal features, combine the characteristics of the Gated recurrent unit with temporal context, and use the Graph Neural Network to predict sensor feature. We tackle the issue of poor sensor network efficiency and sluggish speed without data fusion.

Explainable AI for Healthcare 5.0: Opportunities and Challenges

In the healthcare domain, a transformative shift is envisioned towards Healthcare 5.0. It expands the operational boundaries of Healthcare 4.0 and leverages patient-centric digital wellness. Healthcare 5.0 focuses on real-time patient monitoring, ambient control and wellness, and privacy compliance through assisted technologies like artificial intelligence (AI), Internet-of-Things (IoT), big data, and assisted networking channels. However, healthcare operational procedures, verifiability of prediction models, resilience, and lack of ethical and regulatory frameworks are potential hindrances to the realization of Healthcare 5.0. Recently, explainable AI (EXAI) has been a disruptive trend in AI that focuses on the explainability of traditional AI models by leveraging the decision-making of the models and prediction outputs. The explainability factor opens new opportunities to the black-box models and brings confidence in healthcare stakeholders to interpret the machine learning (ML) and deep learning (DL) models. EXAI is focused on improving clinical health practices and brings transparency to the predictive analysis, which is crucial in the healthcare domain. Recent surveys on EXAI in healthcare have not significantly focused on the data analysis and interpretation of models, which lowers its practical deployment opportunities. Owing to the gap, the proposed survey explicitly details the requirements of EXAI in Healthcare 5.0, the operational and data collection process. Based on the review method and presented research questions, systematically, the article unfolds a proposed architecture that presents an EXAI ensemble on the computerized tomography (CT) image classification and segmentation process. A solution taxonomy of EXAI in Healthcare 5.0 is proposed, and operational challenges are presented. A supported case study on electrocardiogram (ECG) monitoring is presented that preserves the privacy of local models via federated learning (FL) and EXAI for metric validation. The case-study is supported through experimental validation. The analysis proves the efficacy of EXAI in health setups that envisions real-life model deployments in a wide range of clinical applications.

Развитие искусственного интеллекта и глобальный кризис земной цивилизации (к анализу социогуманитарных проблем)

The article considers a qualitatively new stage in the development of artificial intelligence (AI), associated with the development of artificial general intelligence (abbreviated as AGI in the international nomenclature – from Artificial General Intelligence). Unlike traditional AI, AGI is significantly closer in its functions to natural intelligence (EI), it will be able to self-learn, solve a wide range of tasks in different environments, i.e. be integral and autonomous. Such a level of “independence” of AGI opens up fundamentally new prospects for the development of information technologies, but at the same time poses many acute socio-humanitarian problems associated with the risks and threats of losing control over the development of AI. The successful development of AGI requires new theoretical and methodological approaches based on the principles of post-nonclassical epistemology and the results of neuroscientific and phenomenological studies of consciousness. It is very important to consider these issues from the angle of the extreme aggravation of the global crisis of world civilization, due to its consumer dominance and efforts to preserve its monopolar structure from the part of the United States and its Western allies. In this regard, a broader, philosophical-anthropological approach is also required to understand the current state of our civilization and the possibilities for its transformation. It involves taking into account what is called the nature of man, as a stable complex of his mental and bodily properties. They were reproduced among all peoples, in all historical epochs, under all social structures, which indicates their biological conditionality. Among them, along with altruistic properties, a number of negative properties can be distinguished (such as unlimited consumerism, aggressiveness towards one’s own kind, excessive egoistic self-will). These characteristic properties of mass consciousness were actively exploited adherents of monopolarity in their interests. Overcoming the principles and practices of monopolarity and thereby changing the global social self-organization is a necessary condition for a truly humanistic stage of anthropotechnological evolution, capable of opening up new existential prospects for the transformation of man and mankind.

Thermogram classification using deep siamese network for neonatal disease detection with limited data

ABSTRACT Monitoring the body temperatures and evaluating the thermal asymmetry of newborns give an idea about neonatal diseases. Infrared thermography is a non-invasive, non-harmful, and non-contact modality that allows the monitoring of the body temperature distribution. Early diagnosis using a limited data set is extremely vital due to the high mortality rate in newborns and some difficulties in neonatal imaging. Thermography stands out as a useful tool in detecting neonatal diseases compared to other techniques. However, creating a thermogram database consisting of thousands of images from each class required by traditional artificial intelligence methods, is impossible due to the sensitivity of newborns. One of the meta-learning models that has recently gained success in applying limited data learning, especially one-shot, in various fields is Siamese neural networks. In this work, we perform a multi-class classification to provide pre-diagnosis to experts in disease detection using Siamese neural networks. By using two different optimisation techniques and data augmentation, critical diseases with only a few sample data are classified using the method tested in two- and three-class evaluation approaches. The results based on the disease type achieve 99.4% accuracy in infection diseases and 96.4% oesophageal atresia, 97.4% in intestinal atresia, and 94.02% in necrotising enterocolitis.

Financial Risk Early Warning Based on Wireless Network Communication and the Optimal Fuzzy SVM Artificial Intelligence Model

Since the beginning of the new century, risk events such as the world economic crisis have occurred, which have greatly impacted the real economy of our country. A wireless network is a network implemented using wireless communication technology. It includes both global voice and data networks that allow users to establish long-distance wireless connections, as well as infrared technology and radio frequency technology optimized for short-distance wireless connections. These events have a great impact on many small- and medium-sized listed companies, resulting to many small- and medium-sized listed companies going bankrupt. Indeed, one of the important reasons for the frequent bankruptcy of small- and medium-sized listed companies is the lack of awareness of risk prevention and effective financial risk early warning mechanism. The support vector machine is a machine learning method based on the VC dimension theory of statistical learning and the principle of structural risk minimization. This method shows many unique advantages when dealing with classification problems and has been widely used in many fields. The purpose of this article is to realize the financial risk analysis of listed companies through wireless network communication and the optimal fuzzy SVM artificial intelligence model, which help small- and medium-sized listed companies find abnormalities in their business management activities in advance and deal with market risks in a timely manner. Taking 81 small- and medium-sized listed companies as the research object, this paper chooses the small- and medium-sized listed companies in every quarter of 2018 as the research sample. By using the financial and nonfinancial data of small- and medium-sized listed companies and introducing the support vector machine (SVM) with the fuzzy method, the model of the fuzzy support vector machine (FSVM) is constructed. And the performance of the FSVM under four different kernel functions is compared and studied. At the same time, the performance of the FSVM is compared with other artificial intelligence models. The empirical results show that different kernel functions have different effects on the prediction performance of the FCM-SVM model. Under the Gauss radial basis function, the prediction accuracy of the FCM-SVM is over 86%. It can be seen that in predicting the financial crisis of small- and medium-sized listed companies, the FCM-SVM model with Gauss radial basis function has the best predictive performance. The FSVM model based on Gauss radial basis function not only has the advantages of linearity, being polynomial, and nonlinearity of neurons but also is significantly superior to the traditional artificial intelligence model.

Cluster Analysis of Cell Nuclei in H&E-Stained Histological Sections of Prostate Cancer and Classification Based on Traditional and Modern Artificial Intelligence Techniques.

Biomarker identification is very important to differentiate the grade groups in the histopathological sections of prostate cancer (PCa). Assessing the cluster of cell nuclei is essential for pathological investigation. In this study, we present a computer-based method for cluster analyses of cell nuclei and performed traditional (i.e., unsupervised method) and modern (i.e., supervised method) artificial intelligence (AI) techniques for distinguishing the grade groups of PCa. Two datasets on PCa were collected to carry out this research. Histopathology samples were obtained from whole slides stained with hematoxylin and eosin (H&E). In this research, state-of-the-art approaches were proposed for color normalization, cell nuclei segmentation, feature selection, and classification. A traditional minimum spanning tree (MST) algorithm was employed to identify the clusters and better capture the proliferation and community structure of cell nuclei. K-medoids clustering and stacked ensemble machine learning (ML) approaches were used to perform traditional and modern AI-based classification. The binary and multiclass classification was derived to compare the model quality and results between the grades of PCa. Furthermore, a comparative analysis was carried out between traditional and modern AI techniques using different performance metrics (i.e., statistical parameters). Cluster features of the cell nuclei can be useful information for cancer grading. However, further validation of cluster analysis is required to accomplish astounding classification results.

Applying Neuromorphic Computing Simulation in Band Gap Prediction and Chemical Reaction Classification.

The rapidly developing artificial intelligence (AI) requires revolutionary computing architectures to break the energy efficiency bottleneck caused by the traditional von Neumann computing architecture. In addition, the emerging brain–machine interface also requires computational circuitry that can conduct large parallel computational tasks with low energy cost and good biocompatibility. Neuromorphic computing, a novel computational architecture emulating human brains, has drawn significant interest for the aforementioned applications due to its low energy cost, capability to parallelly process large-scale data, and biocompatibility. Most efforts in the domain of neuromorphic computing focus on addressing traditional AI problems, such as handwritten digit recognition and file classification. Here, we demonstrate for the first time that current neuromorphic computing techniques can be used to solve key machine learning questions in cheminformatics. We predict the band gaps of small-molecule organic semiconductors and classify chemical reaction types with a simulated neuromorphic circuitry. Our work can potentially guide the design and fabrication of elementary devices and circuitry for neuromorphic computing specialized for chemical purposes.

Network Security Defense Decision-Making Method Based on Stochastic Game and Deep Reinforcement Learning

Aiming at the existing network attack and defense stochastic game models, most of them are based on the assumption of complete information, which causes the problem of poor applicability of the model. Based on the actual modeling requirements of the network attack and defense process, a network defense decision-making model combining incomplete information stochastic game and deep reinforcement learning is proposed. This model regards the incomplete information of the attacker and the defender as the defender’s uncertainty about the attacker’s type and uses the Double Deep Q-Network algorithm to solve the problem of the difficulty of determining the network state transition probability, so that the network system can dynamically adjust the defense strategy. Finally, a simulation experiment was performed on the proposed model. The results show that, under the same experimental conditions, the proposed method in this paper has a better convergence speed than other methods in solving the defense equilibrium strategy. This model is a fusion of traditional methods and artificial intelligence technology and provides new research ideas for the application of artificial intelligence in the field of cyberspace security.

A fault diagnostic method for oil-immersed transformer based on multiple probabilistic output algorithms and improved DS evidence theory

Power transformer is one of the most vital equipment in the electric power systems, which serves as the connection of power networks at different voltage levels. However, the power transformers in service are subjected to various stresses like thermal stress and electrical stress, which may lead to several faults such as arching, partial discharge and overheating. In order to detect these incipient faults and ensure the normal operation of power transformers, the diagnostic methods based on Dissolved Gas Analysis (DGA) have been developed. However, the DGA methods suffer from low diagnostic accuracy. On the other hand, with the development of Artificial Intelligence (AI), a lot of intelligent algorithms have been applied in this area. However, most traditional AI algorithms are hard classification, which may cause the wrong diagnosis of faults. To overcome those disadvantages, a method based on the improved DS evidence theory and multiple probabilistic output algorithms is proposed in this paper to improve the performance of fault diagnosis of oil-immersed transformers. Firstly, the performance of the soft and hard classification algorithms is compared and analyzed. After that, three diagnostic models based on the Multiclass Relevance Vector Machine (MRVM), Multiclass Support Vector Machine (MSVM) and Back Propagation Neural Network (BPNN) are established. In order to improve the accuracy of these models, Particle Swarm Optimization (PSO) is used to optimize their hyper-parameters. Finally, the combination of these three probabilistic output models is achieved based on the improved Dempster-Shafer (DS) evidence theory, which can help to obtain a more accurate fault diagnosis result by fusing multiple algorithm results. The case study results show that the proposed method achieves a probabilistic output for the fault diagnosis of oil-immersed transformers and overcomes the deficiency of traditional DGA methods which are difficult to get accurate diagnosis results and can’t summarize the fault development rule inductively.

Embedded VR Video Image Control System in the Ideological and Political Teaching System Based on Artificial Intelligence

With the increasing development of multimedia teaching, the combination of virtual reality (VR) and video image control has very attractive development prospects in ideological and political teaching, for example, the use of virtual technology in games and so on. However, most virtual reality environments are currently built, and the functional development of artificial intelligence multimedia teaching systems is not comprehensive. An artificial intelligence VR video image control system is constructed for the multimedia teaching system. This article analyzes the development of artificial intelligence multimedia teaching systems and compares the detection performance and efficiency of traditional methods and artificial intelligence multimedia VR ideological and political teaching. Research shows that, in the use of VR to control the images of ideological and political teaching, the average accuracy of these ten video images is 75.68%. This shows that the video image classification and detection algorithm model based on artificial intelligence in this paper can extract deeper and more abstract features to classify the target. The artificial intelligence VR video image control algorithm constructed in this paper can reduce the maximum failure rate by 49.16%, 61.02%, and 66.94%, respectively. Compared with the traditional algorithm, the artificial intelligence VR video image control algorithm constructed in this paper can reduce the storage access delay time of 10 different video images by an average of 15.93%, can obtain about 9.37% performance optimization, and can reduce the video image control time by 7.28% and 10.63%, respectively. For pictures, the artificial intelligence VR video image control system in this article can increase the performance by up to 28.49%.

Deep Graph Convolutional Networks for Accurate Automatic Road Network Selection

The selection of road networks is very important for cartographic generalization. Traditional artificial intelligence methods have improved selection efficiency but cannot fully extract the spatial features of road networks. However, current selection methods, which are based on the theory of graphs or strokes, have low automaticity and are highly subjective. Graph convolutional networks (GCNs) combine graph theory with neural networks; thus, they can not only extract spatial information but also realize automatic selection. Therefore, in this study, we adopted GCNs for automatic road network selection and transformed the process into one of node classification. In addition, to solve the problem of gradient vanishing in GCNs, we compared and analyzed the results of various GCNs (GraphSAGE and graph attention networks [GAT]) by selecting small-scale road networks under different deep architectures (JK-Nets, ResNet, and DenseNet). Our results indicate that GAT provides better selection of road networks than other models. Additionally, the three abovementioned deep architectures can effectively improve the selection effect of models; JK-Nets demonstrated more improvement with higher accuracy (88.12%) than other methods. Thus, our study shows that GCN is an appropriate tool for road network selection; its application in cartography must be further explored.

Construction design based on particle group optimization algorithm

Abstract The machines exhibit an intelligence which is artificial intelligence (AI), and it is the design of intelligent agents. A system is represented by an intelligent agent who perceives its environment and the success rate is maximized by taking the action. The AI research is highly specialized and there are two subfields and each communication fails often. The popular AI approaches include the traditional symbolic AI and computational intelligence. In order to optimize the seismic design of the reinforced concrete pier structure, the particle swarm optimization (PSO) algorithm and the reaction spectrum analysis method are combined; they establish a regular bridge of the design variable with cross-sectional characteristics and reinforcement ratios, with the target function. The seismic optimization design framework of the pier is transformed into a multi-objective optimization problem. Calculations show that the method can quickly obtain the optimal design parameters that meet multi-objective requirements. The improved PSO main program and the calling push-over program run time are 4.32 and 1347.56 s, respectively; the push-over program running time is 99.68% of the run time of the total program. Optimization of the seismic performance of the rear bridge pier is significantly improved and is more in line with the design method; the design method proposed in this article is more practical.

High-Accuracy Detection of Maize Leaf Diseases CNN Based on Multi-Pathway Activation Function Module

Maize leaf disease detection is an essential project in the maize planting stage. This paper proposes the convolutional neural network optimized by a Multi-Activation Function (MAF) module to detect maize leaf disease, aiming to increase the accuracy of traditional artificial intelligence methods. Since the disease dataset was insufficient, this paper adopts image pre-processing methods to extend and augment the disease samples. This paper uses transfer learning and warm-up method to accelerate the training. As a result, three kinds of maize diseases, including maculopathy, rust, and blight, could be detected efficiently and accurately. The accuracy of the proposed method in the validation set reached 97.41%. This paper carried out a baseline test to verify the effectiveness of the proposed method. First, three groups of CNNs with the best performance were selected. Then, ablation experiments were conducted on five CNNs. The results indicated that the performances of CNNs have been improved by adding the MAF module. In addition, the combination of Sigmoid, ReLU, and Mish showed the best performance on ResNet50. The accuracy can be improved by 2.33%, proving that the model proposed in this paper can be well applied to agricultural production.

Person Re-Identification Microservice over Artificial Intelligence Internet of Things Edge Computing Gateway

With the increase in the number of surveillance cameras being deployed globally, an important topic is person re-identification (Re-ID), which identifies the same person from multiple different angles and different directions across multiple cameras. However, because of the privacy issues involved in the identification of individuals, Re-ID systems cannot send the image data to cloud, and these data must be processed on edge servers. However, there has been a significant increase in computing resources owing to the processing of artificial intelligence (AI) algorithms through edge computing (EC). Consequently, the traditional AI Internet of Things (AIoT) architecture is no longer sufficient. In this study, we designed a Re-ID system at the AIoT EC gateway, which utilizes a microservice to perform Re-ID calculations on EC and balances efficiency with privacy protection. Experimental results indicate that this architecture can provide sufficient Re-ID computing resources to allow the system to scale up or down flexibly to support different scenarios and demand loads.

The Application of Big Data Technology in Computer Artificial Intelligence Technology

Compared with the traditional artificial intelligence design, the artificial intelligence automation design of big data has obvious economic applicability. In recent years, with the continuous development of the field of artificial intelligence design in China, artificial intelligence design has been effectively innovated and promoted, and big data technology, as the representative of which, also plays an obvious role. By expatiating the design, manufacturing and construction concepts of artificial intelligence in the new era, this paper analyzes the principle of big data technology and its deep integration with artificial intelligence.

An interpretable evolving fuzzy neural network based on self-organized direction-aware data partitioning and fuzzy logic neurons

This paper proposes the definition of the architecture of an evolving fuzzy neural network based on self-organizing direction aware data partitioning through stochastic processes based on the dataset used in the model. The choice between four different types of logical neurons in the second layer and the six different types of neuron activation function that compose the artificial neural network of aggregation are defined in a procedure of generating random pairs of combinations between the two factors. The combination that obtains the maximization of training accuracy is chosen to compose the network and perform the composition of the model structure, whose components can be transferred to readable IF-THEN rules for interpretability purposes. Furthermore, the model is able to adapt its parameters and evolve its structure autonomously with new data samples through self-organized direction-aware data partitioning (SODA). In this context, we also propose a technique to measure the degree of changes of neurons (rules), which could be used for structural active learning purposes (e.g, to request user feedback in the case of significant changes). To compare the proposed approach, binary pattern classification tests were performed, and the results were compared with other models of fuzzy neural networks and neural networks, obtaining satisfactory results with the stochastic definition of elements that compose their architecture comparing the final accuracy of the model when classifying real datasets. The proposed model obtained the best result in 3 of the four synthetic bases evaluated, in addition to the best accuracy results in the classification of patterns in five of the nine evaluated real datasets. It highlights the accuracy in problems in patients who underwent breast cancer surgery (72.44%), diabetes evaluation (67.87%), Australian (72.27%) and German (80.51%) credit ratings evaluation, and finally in the classification of radar signals in the ionosphere (90.46%).It is also noteworthy to obtain fuzzy rules in evolution extracted from a real problem of the identification of respiratory diseases through the collection of saliva with 76.67% of accuracy. The results presented by the model were superior to traditional artificial intelligence models, while it was possible to extract knowledge in form of interpretable rules and to realize how these changed over time.

Abstract CT211: Personalized, rational, efficacy-driven chemotherapy dosing via an artificial intelligence system (PRECISE): A protocol for the PRECISE CURATE.AI pilot clinical trial

Abstract Background: Faced with a trade-off between efficacy and toxicity, oncologists have conventionally administered the maximum tolerated doses in chemotherapy on an assumption that higher doses increase efficacy. However, multiple studies have shown that this method of toxicity-guided dosing may result in more frequent toxicities and potentially suboptimal efficacy. With the advent of artificial intelligence (AI), personalized dosing in chemotherapy may be considered to optimize patient care. CURATE.AI is an efficacy-driven, indication-agnostic and mechanism-independent personalized dosing platform that may offer an optimal solution. In contrast to traditional AI approaches based on massive volumes of population data, CURATE.AI requires only the individual patient's medical profile for dose recommendations. Based on the observation that the relationship between a drug dose and a phenotypic response in a human system can be modelled by a dynamic quadratic surface, CURATE.AI continually guides dosing throughout the treatment duration to optimize efficacy. While CURATE.AI has been used in various clinical settings, there are no prior randomized controlled trials (RCTs) on CURATE.AI-guided chemotherapy dosing for solid tumors. Therefore, we are conducting a pilot study to assess the technical and logistical feasibility of an RCT for CURATE.AI-guided solid tumor chemotherapy dosing. We aim to collect exploratory data on efficacy and toxicity, and on the use of longitudinal blood tumor marker measurements, including ctDNA, to inform dose guidance decision. Methods: PRECISE is an open-label, single-arm, multi-centre, prospective pilot clinical trial on using CURATE.AI to achieve personalized, efficacy-driven and dynamically optimized chemotherapy dosing for solid tumors (NCT04522284). Adults with metastatic solid tumors and raised baseline tumor marker levels who are planned for palliative-intent, capecitabine-based chemotherapy will be recruited. CURATE.AI will guide drug dosing for each participant based only on their own tumor marker levels and drug doses as input data. The primary outcome is the proportion of participants in whom CURATE.AI is successfully applied. Secondary outcomes include the timeliness of dose recommendations, participant and physician adherence to CURATE.AI-recommended doses, and the proportion of clinically significant dose changes. As an exploratory outcome, we will analyze the utility of tumor markers including CEA, CA19-9 and ctDNA in high frequency serial measurements. We aim to initially enroll 10 participants from 2 hospitals in Singapore, perform an interim analysis, and consider either cohort expansion or a RCT based on initial pilot data. Recruitment of patients began in August 2020. As of December 2020, 2 participants have been enrolled with recruitment planned for 1 year. Citation Format: Chong Boon Teo, Benjamin Kye Jyn Tan, Xavier Tadeo, Siyu Peng, Hazel Pei Lin Soh, Sherry De Xuan Du, Vilianty Wen Ya Luo, Aishwarya Bandla, Raghav Sundar, Dean Ho, Theodore Kee, Agata Blasiak. Personalized, rational, efficacy-driven chemotherapy dosing via an artificial intelligence system (PRECISE): A protocol for the PRECISE CURATE.AI pilot clinical trial [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr CT211.

Predictive intelligence powered attentional stacking matrix factorization algorithm for the computational drug repositioning

Artificial Intelligence (AI) technologies are widely used to study the computational drug repositioning to find potential new uses for marketed drugs, which have a huge impact on drug development. However, due to the sparsity of datasets and the linear limitations of traditional AI or machine learning models lead to the inability of most methods to effectively obtain hidden feature of drugs and diseases To end this, Predictive Intelligence (PI) has been proposed in recent years. Hence, in this work, we propose a PI powered Attentional Stacking Matrix Factorization (PI-ASMF). Firstly, in order to extract the effective latent factor of the drugs or diseases the ASMF model superimposes the drug–disease similarity information and the drug–disease association information into the PI model, which ensures that it can learn the effective latent factor and alleviates the cold start problem to a certain extent. Subsequently, the latent factors of drugs and diseases are element-wise multiplied with their respective attentional vectors to learn the adaptive weights of each feature. Finally, the latent factors of drugs and diseases are fed into the prediction module to generate predictive values of potential drug–disease associations. The AUC and AUPR values of PI-ASMF model on two real datasets are 0.892 and 0.168, 0.913 and 0.25, respectively, which verified its superiority and validity.