Pub Date : 2022-03-30DOI: 10.1109/ICAI55435.2022.9773637
M. Talha, Sheikh Faisal Rashid, Zain Iftikhar, Muhammad Touseef Afzal, Liu Ying
In recent years, convolutional neural networks (CNNs) have become a de facto standard in computer vision for object detection and recognition. At present, CNNs have been used in many application areas including the automation of industrial manufacturing processes. But using CNN in a real-time environment to track defects on products has many shortcomings like long training time, large data requirements, slow inference time, dynamic environment, and hardware dependency. This paper evaluates the state-of-the-art CNN architectures for object detection to address the mentioned challenges and provide the best possible solution. A set of pre-trained models has been trained on just 781 annotated images by applying transfer learning. Experimental results showed that Faster RCNN with VGG-16 backbone outperforms the other models in case of accuracy and mAP. But RetinaNet with an FPN backbone has the fastest inference time on multi-scaled defects. Paper also presents the deployment pipeline for inference on mobile devices to use in a real-time environment without any special hardware. In addition, an improved dataset of submersible pump impellers, based on the existing Kaggle dataset is introduced.
{"title":"Transferable Learning Architecture for Scalable Visual Quality Inspection","authors":"M. Talha, Sheikh Faisal Rashid, Zain Iftikhar, Muhammad Touseef Afzal, Liu Ying","doi":"10.1109/ICAI55435.2022.9773637","DOIUrl":"https://doi.org/10.1109/ICAI55435.2022.9773637","url":null,"abstract":"In recent years, convolutional neural networks (CNNs) have become a de facto standard in computer vision for object detection and recognition. At present, CNNs have been used in many application areas including the automation of industrial manufacturing processes. But using CNN in a real-time environment to track defects on products has many shortcomings like long training time, large data requirements, slow inference time, dynamic environment, and hardware dependency. This paper evaluates the state-of-the-art CNN architectures for object detection to address the mentioned challenges and provide the best possible solution. A set of pre-trained models has been trained on just 781 annotated images by applying transfer learning. Experimental results showed that Faster RCNN with VGG-16 backbone outperforms the other models in case of accuracy and mAP. But RetinaNet with an FPN backbone has the fastest inference time on multi-scaled defects. Paper also presents the deployment pipeline for inference on mobile devices to use in a real-time environment without any special hardware. In addition, an improved dataset of submersible pump impellers, based on the existing Kaggle dataset is introduced.","PeriodicalId":146842,"journal":{"name":"2022 2nd International Conference on Artificial Intelligence (ICAI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133773444","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-30DOI: 10.1109/ICAI55435.2022.9773481
Musa Khan, G. M. Khan
Predicting the Covid-19 spread and its impact on the stock market is an important research challenge these days. In order to obtain the best forecasting model, we have exploited neuro-evolutionary technique Cartesian genetic programming evolved artificial neural network (CGPANN) based solution to predict the future cases of COVID-19 up to 6-days in advance. This helps authorities and paramedical staff to take precautionary measures on time which helps in counteracting the spreading of the virus. The rising number of COVID cases has caused a significant impact on the stock market. CGPANN being the best performer for the time series prediction model seems ideal for the case under consideration. The proposed model achieved an accuracy as high as 98% predicting COVID-19 cases for the next six days. When compared with other contemporary models CGPANN seems to perform well ahead in terms of accuracy.
{"title":"COVID-19 Spread Prediction and Its Impact on the Stock market price","authors":"Musa Khan, G. M. Khan","doi":"10.1109/ICAI55435.2022.9773481","DOIUrl":"https://doi.org/10.1109/ICAI55435.2022.9773481","url":null,"abstract":"Predicting the Covid-19 spread and its impact on the stock market is an important research challenge these days. In order to obtain the best forecasting model, we have exploited neuro-evolutionary technique Cartesian genetic programming evolved artificial neural network (CGPANN) based solution to predict the future cases of COVID-19 up to 6-days in advance. This helps authorities and paramedical staff to take precautionary measures on time which helps in counteracting the spreading of the virus. The rising number of COVID cases has caused a significant impact on the stock market. CGPANN being the best performer for the time series prediction model seems ideal for the case under consideration. The proposed model achieved an accuracy as high as 98% predicting COVID-19 cases for the next six days. When compared with other contemporary models CGPANN seems to perform well ahead in terms of accuracy.","PeriodicalId":146842,"journal":{"name":"2022 2nd International Conference on Artificial Intelligence (ICAI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133394080","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-30DOI: 10.1109/ICAI55435.2022.9773418
Kainat, Sara Ali, Fahad Iqbal Khawaja, Yasar Avaz, Muhammad Saiid
Analyzing one's participation and attention may be useful in a variety of contexts, like work situations such as driving a car, defusing a bomb, and many learning environments. Increasing the student's involvement and participation in the classroom has been proven to improve learning results. Attention is core for effective learning, yet analyzing attention is a tricky task. People have been working on attention analysis for decades, and as a result, current learning systems contain methods for monitoring and reporting on students' attention states. Facial features and eye movements are some of the important behavioural features to access attentiveness. Approaches such as EEG signals, gaze detection, head and body posture detection are used in this context as they provide rich information about a person's behavior and thoughts. It also gives essential information for interpreting their nonverbal, cues. These are referred to be “honest signals” since they are unconscious patterns that reveal the focus of our attention. They give vital indications concerning teaching methods and students' responses to various conscious and unconscious teaching tactics inside the classroom. Examining verbal and nonverbal conduct in the classroom can give valuable input to the instructor. This paper will go through various approaches available for analyzing student attentiveness for effective learning in the classroom. Integrating different technical approaches with Machine learning and Deep learning models accuracy up to 90% can be observed in different research with minimum error.
{"title":"A Review on Different Approaches for Assessing Student Attentiveness in Classroom using Behavioural Elements","authors":"Kainat, Sara Ali, Fahad Iqbal Khawaja, Yasar Avaz, Muhammad Saiid","doi":"10.1109/ICAI55435.2022.9773418","DOIUrl":"https://doi.org/10.1109/ICAI55435.2022.9773418","url":null,"abstract":"Analyzing one's participation and attention may be useful in a variety of contexts, like work situations such as driving a car, defusing a bomb, and many learning environments. Increasing the student's involvement and participation in the classroom has been proven to improve learning results. Attention is core for effective learning, yet analyzing attention is a tricky task. People have been working on attention analysis for decades, and as a result, current learning systems contain methods for monitoring and reporting on students' attention states. Facial features and eye movements are some of the important behavioural features to access attentiveness. Approaches such as EEG signals, gaze detection, head and body posture detection are used in this context as they provide rich information about a person's behavior and thoughts. It also gives essential information for interpreting their nonverbal, cues. These are referred to be “honest signals” since they are unconscious patterns that reveal the focus of our attention. They give vital indications concerning teaching methods and students' responses to various conscious and unconscious teaching tactics inside the classroom. Examining verbal and nonverbal conduct in the classroom can give valuable input to the instructor. This paper will go through various approaches available for analyzing student attentiveness for effective learning in the classroom. Integrating different technical approaches with Machine learning and Deep learning models accuracy up to 90% can be observed in different research with minimum error.","PeriodicalId":146842,"journal":{"name":"2022 2nd International Conference on Artificial Intelligence (ICAI)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116282760","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-30DOI: 10.1109/ICAI55435.2022.9773489
Ihsan UI Haq, R. Mumtaz, M. Talha, Zunaira Shafaq, M. Owais
Wheat leaf rust is considered one of the most detrimental fungal diseases that spread rapidly after its first appearance and can significantly damage the entire crop field. This can lead to a severe decline in wheat yield, posing a serious threat to food security considering an unceasing growth in the country's population. The conventional method of wheat rust detection is visual inspection, which is an ineffective and unsuitable approach for large agricultural lands. Additionally, such monitoring is solely dependent on the farmer's knowledge base and experience. Towards such an end, an Edge AI-based system for detecting and classifying wheat leaves into healthy and rusted leaves in real-time is proposed. The dataset collected is analyzed with several machine learning-based classifiers where Random Forest outperformed with a classification accuracy of 97.3% and 82.8% using Gray Level Co-occurrence Matrix (GLCM) and binary feature extraction techniques respectively. In addition, a Deep Convolution Neural Network (DCNN) model is explored to classify rusted and healthy leaves, which showed an accuracy of 88.33 %. This trained DCNN model is also deployed on the edge device for real-time classification of wheat rust disease. The developed system would contribute to promoting technology-based solutions over old farming practices and assist in minimizing the spread of wheat rust disease.
{"title":"Wheat Rust Disease Classification using Edge-AI","authors":"Ihsan UI Haq, R. Mumtaz, M. Talha, Zunaira Shafaq, M. Owais","doi":"10.1109/ICAI55435.2022.9773489","DOIUrl":"https://doi.org/10.1109/ICAI55435.2022.9773489","url":null,"abstract":"Wheat leaf rust is considered one of the most detrimental fungal diseases that spread rapidly after its first appearance and can significantly damage the entire crop field. This can lead to a severe decline in wheat yield, posing a serious threat to food security considering an unceasing growth in the country's population. The conventional method of wheat rust detection is visual inspection, which is an ineffective and unsuitable approach for large agricultural lands. Additionally, such monitoring is solely dependent on the farmer's knowledge base and experience. Towards such an end, an Edge AI-based system for detecting and classifying wheat leaves into healthy and rusted leaves in real-time is proposed. The dataset collected is analyzed with several machine learning-based classifiers where Random Forest outperformed with a classification accuracy of 97.3% and 82.8% using Gray Level Co-occurrence Matrix (GLCM) and binary feature extraction techniques respectively. In addition, a Deep Convolution Neural Network (DCNN) model is explored to classify rusted and healthy leaves, which showed an accuracy of 88.33 %. This trained DCNN model is also deployed on the edge device for real-time classification of wheat rust disease. The developed system would contribute to promoting technology-based solutions over old farming practices and assist in minimizing the spread of wheat rust disease.","PeriodicalId":146842,"journal":{"name":"2022 2nd International Conference on Artificial Intelligence (ICAI)","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126100245","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-30DOI: 10.1109/ICAI55435.2022.9773663
Maheen Ayub, N. A. Khan, R. Haider
Agriculture plays an important role in the growth of a country's economy. Crop area and yield predictions using machine learning are important investigation domains in current research fields. Wheat is the most important food crop in Pakistan which is cultivated in the Rabi season. Weather conditions, Remote Sensing (RS) data, and Machine learning (ML) technologies can be used to forecast wheat yield before actual harvesting to assist the management of wheat production, trade, and storage. In this paper, a supervised ML based framework is proposed that extracts features/Vegetation Indices (VIs) including Enhanced Vegetation Index (EVI), Normalized Difference Vegetation Index (NDVI), Red Edge Normalized Difference Vegetation Index (RENDVI), and Normalized Difference Moisture Index (NDMI) from Sentinel-2 Satellite images and contributes for: estimation of wheat area, and identification of most effective VIs in wheat area estimation, prediction of wheat yield, and identification of most effective VIs and meteorological parameters in wheat yield prediction. In the initial experimental setup, good performance output obtained using the Random Forest (RF) machine learning algorithm therefore in this framework RF machine learning algorithm is focused on wheat area estimation and generation of Land Use Land Cover (LULC) maps which is capable of estimating area with an accuracy of 84%, consumer's accuracy of 81 %, producer's accuracy of 83% and kappa statistics of 0.80. LULC maps are used for wheat yield prediction. Multivariate regression forward stepwise technique is applied for yield prediction and selection of effective VIs and meteorological parameters. The adjusted coefficient of determination (R2) between reported and predicted yield found 0.84 with an error of 46.14 Kg/ha for yield prediction.
{"title":"Wheat Crop Field and Yield Prediction using Remote Sensing and Machine Learning","authors":"Maheen Ayub, N. A. Khan, R. Haider","doi":"10.1109/ICAI55435.2022.9773663","DOIUrl":"https://doi.org/10.1109/ICAI55435.2022.9773663","url":null,"abstract":"Agriculture plays an important role in the growth of a country's economy. Crop area and yield predictions using machine learning are important investigation domains in current research fields. Wheat is the most important food crop in Pakistan which is cultivated in the Rabi season. Weather conditions, Remote Sensing (RS) data, and Machine learning (ML) technologies can be used to forecast wheat yield before actual harvesting to assist the management of wheat production, trade, and storage. In this paper, a supervised ML based framework is proposed that extracts features/Vegetation Indices (VIs) including Enhanced Vegetation Index (EVI), Normalized Difference Vegetation Index (NDVI), Red Edge Normalized Difference Vegetation Index (RENDVI), and Normalized Difference Moisture Index (NDMI) from Sentinel-2 Satellite images and contributes for: estimation of wheat area, and identification of most effective VIs in wheat area estimation, prediction of wheat yield, and identification of most effective VIs and meteorological parameters in wheat yield prediction. In the initial experimental setup, good performance output obtained using the Random Forest (RF) machine learning algorithm therefore in this framework RF machine learning algorithm is focused on wheat area estimation and generation of Land Use Land Cover (LULC) maps which is capable of estimating area with an accuracy of 84%, consumer's accuracy of 81 %, producer's accuracy of 83% and kappa statistics of 0.80. LULC maps are used for wheat yield prediction. Multivariate regression forward stepwise technique is applied for yield prediction and selection of effective VIs and meteorological parameters. The adjusted coefficient of determination (R2) between reported and predicted yield found 0.84 with an error of 46.14 Kg/ha for yield prediction.","PeriodicalId":146842,"journal":{"name":"2022 2nd International Conference on Artificial Intelligence (ICAI)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122453662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-30DOI: 10.1109/ICAI55435.2022.9773706
Niaz Muhammad, Syed Waqar Shah, G. M. Khan
Financial market volatility has become one of the most difficult applications for stock price forecasting in ongoing situations. The current statistical models for stock price forecasting are too rigid and inefficient to appropriately deal with the uncertainty and volatility inherent in stock data. CGPANN-CGP based ANNs and LSTM are the most common methods used these days to predict such dynamics in time series data. In comparison to other methodologies, studies have demonstrated that the application of Cartesian genetic programming evolved Artificial Neural Networks (CGPANNs) to time series forecasting problems produces better results, and LSTM can be competitive at times. CGPANN provides the ability to train both structure, topology, and weights of network to achieve the global optimum solution. The prediction model is trained on the behavior of stock exchange patterns and is based on trends in historical daily stock prices. The proposed CGPANN and LSTM models produced competitive results of 98.86% and 98.52% respectively. However, CGPANN architecture is capable computationally efficient than LSTM and its ability of quick predictions makes it ideal for real-time applications.
{"title":"Evolving computationally efficient prediction model for Stock Volatility using CGPANN","authors":"Niaz Muhammad, Syed Waqar Shah, G. M. Khan","doi":"10.1109/ICAI55435.2022.9773706","DOIUrl":"https://doi.org/10.1109/ICAI55435.2022.9773706","url":null,"abstract":"Financial market volatility has become one of the most difficult applications for stock price forecasting in ongoing situations. The current statistical models for stock price forecasting are too rigid and inefficient to appropriately deal with the uncertainty and volatility inherent in stock data. CGPANN-CGP based ANNs and LSTM are the most common methods used these days to predict such dynamics in time series data. In comparison to other methodologies, studies have demonstrated that the application of Cartesian genetic programming evolved Artificial Neural Networks (CGPANNs) to time series forecasting problems produces better results, and LSTM can be competitive at times. CGPANN provides the ability to train both structure, topology, and weights of network to achieve the global optimum solution. The prediction model is trained on the behavior of stock exchange patterns and is based on trends in historical daily stock prices. The proposed CGPANN and LSTM models produced competitive results of 98.86% and 98.52% respectively. However, CGPANN architecture is capable computationally efficient than LSTM and its ability of quick predictions makes it ideal for real-time applications.","PeriodicalId":146842,"journal":{"name":"2022 2nd International Conference on Artificial Intelligence (ICAI)","volume":"90 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116002570","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-30DOI: 10.1109/ICAI55435.2022.9773697
Saad A. Bazaz, AbdurRehman Subhani, Syed Z.A. Hadi
With the recent global boom in video content creation and consumption during the pandemic, linguistics remains the only barrier in producing im-mersive content for global communities. To solve this, content creators use a manual dubbing process, where voice actors are hired to produce a “voiceover” over the video. We aim to break down the language barrier and thus make “videos for everyone”. We propose an end-to-end architecture that automatically translates videos and produces synchronized dubbed voices using deep learning models, in a specified target language. Our architecture takes a modular approach, allowing the user to tweak each component or replace it with a better one. We present our results from said architecture, and describe possible future motivations to scale this to accommodate multiple languages and multiple use cases. A sample of our results can be found here: https://youtu.be/eGB-gL6bDr4
{"title":"Automated Dubbing and Facial Synchronization using Deep Learning","authors":"Saad A. Bazaz, AbdurRehman Subhani, Syed Z.A. Hadi","doi":"10.1109/ICAI55435.2022.9773697","DOIUrl":"https://doi.org/10.1109/ICAI55435.2022.9773697","url":null,"abstract":"With the recent global boom in video content creation and consumption during the pandemic, linguistics remains the only barrier in producing im-mersive content for global communities. To solve this, content creators use a manual dubbing process, where voice actors are hired to produce a “voiceover” over the video. We aim to break down the language barrier and thus make “videos for everyone”. We propose an end-to-end architecture that automatically translates videos and produces synchronized dubbed voices using deep learning models, in a specified target language. Our architecture takes a modular approach, allowing the user to tweak each component or replace it with a better one. We present our results from said architecture, and describe possible future motivations to scale this to accommodate multiple languages and multiple use cases. A sample of our results can be found here: https://youtu.be/eGB-gL6bDr4","PeriodicalId":146842,"journal":{"name":"2022 2nd International Conference on Artificial Intelligence (ICAI)","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127274927","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
One of the Challenging tasks in medical field and computer vision is automatic brain segmentation with MRI (Magnetic Resonance Images). From the literature, the importance of deep neural networks is cleared as they have provided effective results in brain tumor segmentation problem in terms of accuracy and time. Mostly the training time is issued due to image features and for this purpose extra computational power is required to train the neural network model. The gradient problem is overcome in this study to fine tune the Novel unit model. CNNs & U-Shaped encoder decoder architectures produce effective result than other neural networks in terms of accuracy and time. The comparison is also performed in this study to show the robustness of U- Shaped encoder decoder architecture. Novel encoder and decoder model accuracy is 0.947 %that is better than other neural networks e.g., CNNs. Further this model is roughly three time faster than other models in terms of training time that's why less computation power is required to train this model.
{"title":"Segmentation of Brain Tumor from Medical Images with Novel U-Shaped Encoder Decoder Architecture","authors":"Farzana Mushtaq, Faisal Rehman, Hira Akram, Sameen Butt, Syeda Fareeha Batool, Maheen Jafer, Nadeem Sarfaraz, Anza Gul","doi":"10.1109/ICAI55435.2022.9773383","DOIUrl":"https://doi.org/10.1109/ICAI55435.2022.9773383","url":null,"abstract":"One of the Challenging tasks in medical field and computer vision is automatic brain segmentation with MRI (Magnetic Resonance Images). From the literature, the importance of deep neural networks is cleared as they have provided effective results in brain tumor segmentation problem in terms of accuracy and time. Mostly the training time is issued due to image features and for this purpose extra computational power is required to train the neural network model. The gradient problem is overcome in this study to fine tune the Novel unit model. CNNs & U-Shaped encoder decoder architectures produce effective result than other neural networks in terms of accuracy and time. The comparison is also performed in this study to show the robustness of U- Shaped encoder decoder architecture. Novel encoder and decoder model accuracy is 0.947 %that is better than other neural networks e.g., CNNs. Further this model is roughly three time faster than other models in terms of training time that's why less computation power is required to train this model.","PeriodicalId":146842,"journal":{"name":"2022 2nd International Conference on Artificial Intelligence (ICAI)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124480940","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-30DOI: 10.1109/ICAI55435.2022.9773495
Touqir Gohar, L. Hasan, G. M. Khan, Mehreen Mubashir
Floods are the world's most damaging natural disasters, which not only claim thousands of human lives but also result in huge damage to infrastructure. Floods if forecasted in advance can help in the reduction of damages. Flood prediction especially long term is a complex task as it involves many hydrological and metrological parameters. For the short and medium-term, machine learning methods seem to have contributed to a great extent in simulating mathematical modelling of the physical flow processes of floods. However, these developed model's performance lacks generalization. Such systems trained on one geographical location's data have degraded performance when exploited for another location. In this paper, Long Short-Term Memory (LSTM) machine learning algorithm was applied where the hourly river level, river flow, and rainfall data from Brooklyn station was used as input data to the model and test for one hour, two hours, four hours, six hours, eight hours, and twelve hours in advance for river level prediction at Hoppers Crossing station. The developed algorithm achieved an accuracy of 98% for one hour and 97.2 %, 96.14 %, 94.67%,94.61 %, and 93.55% for two, four, six, eight, and twelve hours respectively. These systems not only forecast the future water level but also help in estimating the water level in case of a sensor failure. Multivariate modelling is utilized to predict the unknown parameter from the given other parametric values, thus not only predicting the forecasted water level but also reporting the sensor failure.
{"title":"Constraint Free Early Warning System for Flood Using Multivariate LSTM Network","authors":"Touqir Gohar, L. Hasan, G. M. Khan, Mehreen Mubashir","doi":"10.1109/ICAI55435.2022.9773495","DOIUrl":"https://doi.org/10.1109/ICAI55435.2022.9773495","url":null,"abstract":"Floods are the world's most damaging natural disasters, which not only claim thousands of human lives but also result in huge damage to infrastructure. Floods if forecasted in advance can help in the reduction of damages. Flood prediction especially long term is a complex task as it involves many hydrological and metrological parameters. For the short and medium-term, machine learning methods seem to have contributed to a great extent in simulating mathematical modelling of the physical flow processes of floods. However, these developed model's performance lacks generalization. Such systems trained on one geographical location's data have degraded performance when exploited for another location. In this paper, Long Short-Term Memory (LSTM) machine learning algorithm was applied where the hourly river level, river flow, and rainfall data from Brooklyn station was used as input data to the model and test for one hour, two hours, four hours, six hours, eight hours, and twelve hours in advance for river level prediction at Hoppers Crossing station. The developed algorithm achieved an accuracy of 98% for one hour and 97.2 %, 96.14 %, 94.67%,94.61 %, and 93.55% for two, four, six, eight, and twelve hours respectively. These systems not only forecast the future water level but also help in estimating the water level in case of a sensor failure. Multivariate modelling is utilized to predict the unknown parameter from the given other parametric values, thus not only predicting the forecasted water level but also reporting the sensor failure.","PeriodicalId":146842,"journal":{"name":"2022 2nd International Conference on Artificial Intelligence (ICAI)","volume":"48 8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132994465","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-30DOI: 10.1109/ICAI55435.2022.9773590
Ali Bin Mazhar Sultan, S. Mehmood, Hamza Zahid
The usage of appropriate data communication protocols is critical for long-term Internet of Things (IoT) implementation and operation. The publish/subscribe-based Message Queuing Telemetry Transport (MQTT) protocol is widely used in the IoT world. Cyber threats on devices and networks using MQTT protocols are expected to rise with the protocol's growing popularity among IoT manufacturers. Among these threats is the man in the middle (MiTM) threat, in which an attacker listens in on or modifies traffic between two parties by intercepting conversations between them. In this paper we have implemented five different machine learning model on an open-source dataset and evaluated different parameters like accuracy, precision, recall, F1 score and most importantly training time and test time because most of IoT network are hosted on resource constrained devices like Raspberry Pi.
{"title":"Man in the Middle Attack Detection for MQTT based IoT devices using different Machine Learning Algorithms","authors":"Ali Bin Mazhar Sultan, S. Mehmood, Hamza Zahid","doi":"10.1109/ICAI55435.2022.9773590","DOIUrl":"https://doi.org/10.1109/ICAI55435.2022.9773590","url":null,"abstract":"The usage of appropriate data communication protocols is critical for long-term Internet of Things (IoT) implementation and operation. The publish/subscribe-based Message Queuing Telemetry Transport (MQTT) protocol is widely used in the IoT world. Cyber threats on devices and networks using MQTT protocols are expected to rise with the protocol's growing popularity among IoT manufacturers. Among these threats is the man in the middle (MiTM) threat, in which an attacker listens in on or modifies traffic between two parties by intercepting conversations between them. In this paper we have implemented five different machine learning model on an open-source dataset and evaluated different parameters like accuracy, precision, recall, F1 score and most importantly training time and test time because most of IoT network are hosted on resource constrained devices like Raspberry Pi.","PeriodicalId":146842,"journal":{"name":"2022 2nd International Conference on Artificial Intelligence (ICAI)","volume":"302 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121266329","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：