Google Colab Research Articles

Python script fuzzy time series Markov chain model for forecasting the number of diseases cocoa plant in Bendungan district

Cocoa is a plantation commodity whose role is essential for the economy, so it is necessary to be aware of its health to maximize production. Cocoa plant disease data is a time series data because it is observed continuously. One of the time series forecasting models is the Fuzzy Time Series Markov Chain (FTS-MC), a combination of the Fuzzy Time Series (FTS) and Markov Chain models. The model uses the principle of fuzzy logic by transferring FTS data to fuzzy logic and using the obtained fuzzy logic groups to derive the Markov chain transition matrix. In this research, a Python script of the FTS-MC model was built in the Google Colaboratory to forecast the number of cocoa plant diseases in Bendungan District to simplify and speed up the data processing. Python was used in this research because of its easy-to-use, flexible, and open-source syntax. In building Python scripts, libraries and functions are needed by utilizing loop processes and if-else statements. Based on the processing results, forecasting with the FTS-MC model using Python only takes less than 1 minute.

Dimensional Reduction of Underwater Shrimp Digital Image Using the Principal Component Analysis Algorithm

Shrimps are aquaculture products highly needed by the people and this is the reason their growth needs to be monitored using underwater digital images. However, the large dimensions of the shrimp digital images usually make the processing difficult. Therefore, this research focuses on reducing the dimensions of underwater shrimp digital images without reducing their information through the application of the Principal Component Analysis (PCA) algorithm. This was achieved using 4 digital shrimp images extracted from video data with the number of columns 398 for each image. The results showed that 12 PCs were produced and this means the reduced digital images with new dimensions have 12 variable columns with data diversity distributed based on a total variance of 95.61%. Moreover, the original and reduced digital images were compared and the lowest value of MSE produced was 94.12, the minimum value of RMSE was 9.54, and the highest value of PSNR was 8.06 db, and they were obtained in the 4th digital image. The experiment was conducted using 3 devices which include I3, I7, and Google Colab processor computers and the fastest computational result was produced at 2.1 seconds by the Google Colab processor. This means the PCA algorithm is good for the reduction of digital image dimensions as indicated by the production of 12 PC as the new variable dimensions for the reduced underwater image of shrimps.

Use of data mining to establish associations between Indian marine fish catch and environmental data

For decades, changes in fish catch composition and the marine environment have been monitored worldwide and recorded in databases like FAO FishStatJ and the European Union Copernicus Marine Service. However, the complexity and high variability in the dataset makes it challenging to find meaningful information through conventional data analytical methods. Therefore, in this pilot data mining study, we employed association rule mining algorithms (Apriori, ECLAT, and FP-Growth) to find frequently occurring itemsets in the fish-catch composition and marine environment data of the west and east coasts of India during the past decade (2011-2020). Firstly, the inherent spatial and temporal variations in fish-catch composition and marine environment (sea surface temperature and chlorophyll) on the west and east coasts of India were statistically analyzed and described. Then, the data were preprocessed, selected, and transformed into categorical attributes. By applying the association rule mining algorithms written in the Python language in the Google Colab workspace, we obtained frequent itemsets of fish catch and marine environment with different levels of minimum support and confidence. The preliminary results showed linear and inverse associations between changes in the sea surface temperature, chlorophyll concentration, and major catch groups, such as anchovies, Indian oil sardine, Indian mackerel, hairtails, butterfish-pomfrets, Bombay duck, flatfish, tunas, giant tiger prawn, crabs, lobsters, and cephalopods. Among the tested data mining algorithms, FP-Growth was found to be more efficient and reliable in finding associations between the spatiotemporal dynamics of the marine environment and fish distribution and abundance. Therefore, it can be potentially used to support marine fisheries? resource assessment and management strategies after refinement.

Probing antibacterial drugs for Fusobacterium nucleatum subsp. nucleatum ATCC 25586 targeting UDP-N-acetylglucosamine 1-carboxyltransferase.

Fusobacterium nucleatum is a Gram-negative anaerobic bacteria that is commonly found in oral cavities and is associated with connective tissue destruction in periodontitis. UDP-N-acetylglucosamine 1-carboxyltransferase with enzyme commission number 2.5.1.7 is a transferases enzyme that plays a role in bacterial pathogenesis. Inhibiting binding sites of UDP-N-acetylglucosamine 1-carboxyltransferase is needed to find potential antibiotic candidates for periodontitis treatment. Hence, the research aimed to present potential UDP-N-acetylglucosamine 1-carboxyltransferase inhibiting compounds through molecular docking simulation by in silico analysis. DrugBank database was used to obtain the antibacterial candidates, which were further screened computationally using the AutoDock Vina program on Google Colab Pro. The top nine compounds yielded binding affinity ranging from -12.1 to -12.8 kcal/mol, with conivaptan as one of the three compounds having the highest binding affinity. Molecular dynamic study revealed that the ligand-protein complex for conivaptan had root-mean-square deviation values of 0.05-1.1 nm, indicating likeliness for stable interaction. Our findings suggest that conivaptan is the potent UDP-N-acetylglucosamine 1-carboxyltransferase inhibitor, hence its efficacy against periodontitis-causing bacteria.

IMPROVED TRAFFIC CONTROL SYSTEM THROUGH DETECTION AND CLASSIFICATION OF VEHICLES USING CAMERA BASED CROWD SENSING WITH DEEP LEARNING APPROACH.

Abstract - Intelligent traffic light control system can be considered as one of the main aspects of a smart city. There is significant increase in the number of People that own vehicles and this leads to traffic congestion in urban cities of developing countries. Traffic congestion is one of the highly challenging problems people face in their daily lives which affect their working productivity, causes accident, unnecessary delays and possibly loss of lives. The control of traffics based on the densities and classes of the vehicles using deep learning approach proved to be a better approach in handling traffic jams according to the current state of the studies. However, the current state of the studies did not have special consideration for the class of special vehicles (VIPs) which need special consideration due to their importance and tight schedules. Also, it did not consider allocating time based on the classification. Thus, a model to improve the detection and classification of vehicles on side of intersection as well as special consideration for VIPs was proposed. The proposed model was built using Faster R-CNN based on Tensorflow machine learning framework on a Google Colab environment. The model was able to detect and classify vehicles as “special vehicles (i.e. VIP), Emergency (EV) and Others” classes of vehicles, a time allocation simulator was proposed based on the specified classes with special timing for VIPs. Results show that using the proposed method, 94.43% accuracy was achieved.

КВАНТОВОЕ МАШИННОЕ ОБУЧЕНИЕ НА БАЗЕ БИБЛИОТЕК PENNYLANE НА ПРИМЕРЕ ЭКОНОМЕТРИЧЕСКОГО МОДЕЛИРОВАНИЯ КОТИРОВОК ВАЛЮТНОЙ ПАРЫ USD/RUB

The article is devoted to econometric calculations using quantum machine learning technologies and PennyLane libraries. The features of Penny Lane integration into the Google Colab development environment are considered. An econometric model has been created to predict the behavior of the USD/RUB currency pair in order to determine its satisfaction with the values laid down in the framework of the execution of the government budget.

RFSE-GRU: Data Balanced Classification Model for Mobile Encrypted Traffic in Big Data Environment

With the widespread use of mobile technologies and the Internet, traffic in mobile networks is increasing. This situation has made the classification of traffic an important element for data security and network management. However, encryption of traffic in modern networks makes it difficult to classify traffic with traditional methods. In this study, a unique deep learning-based classification model is proposed for the classification of encrypted mobile traffic data. The proposed model is a classification model called RFSE-GRU, which combines the Gated Recurrent Units (GRU) algorithm, feature selection and data balancing. The features that are more meaningful in the classification process are determined by selecting the features with the Random Forest algorithm. In addition, Synthetic Minority Oversampling Technique (SMOTE) oversampling algorithm and Edited Nearest Neighbor (ENN) undersampling algorithm were used together to reduce the negative impact of data imbalance on classification performance. The study was carried out on Apache Spark big data platform in Google Colab environment. In the study, ISCX VPN-Non VPN and UTMobileNet2021 datasets were used. Binary and multiclass classifications were made for the ISCX VPN-Non VPN dataset, and multiclass classifications were made for the UTMobileNet2021 dataset by using various algorithms on the datasets. The proposed model has been compared with eleven different algorithms and hybrid methods. At the same time, the effect of data balancing and feature selection on classification performance is examined. As a result, the proposed model achieved 93.91%, 82.68% and 96.83% accuracy rates in ISCX VPN-Non VPN binary and multiclass, UTMobileNet2021 multiclass classifications, respectively.

Convolutional Neural Network (CNN) Method for Classification of Images by Age

Image classification is one of the studies that is currently being developed. The details of the characteristics that must be captured make researchers compete to find the most suitable method for classifying. The Convolutional Neural Network (CNN) algorithm is one of the most superior algorithms in the field of object classification and identification today. With the help of several packages contained in Google Colab for classification, this algorithm is easier to use. In this study, the target case is the age of a person who will be classified using photos or images taken from the internet which are then stored in the form of Google Drive. The research data used is divided into 2 parts, namely for training data as many as 23.440 images, and 10,046 for testing data. Then to facilitate the extraction of features from the features to be identified, the researchers carried out the preprocessing stage, namely grayscale images, and data augmentation. The purpose of this study is to implement the concept of Deep Learning with Convolutional Neural Networks (CNN) in image classification and to determine the level of accuracy of the CNN model in classifying images. After the algorithm is run and the model has been formed, an accuracy of 78.5% is obtained. It can be concluded that the Convolutional Neural Network (CNN) method is good at classifying images

Comparison Study of Convolutional Neural Network Architecture in Aglaonema Classification

Convolutional Neural Network (CNN) is very good at classifying images. To measure the best CNN architecture, a study must be done against real-case scenarios. Aglaonema, one of the plants with high similarity, is chosen as a test case to compare CNN architecture. In this study, a classification process was carried out on five classes of Aglaonema imagery by comparing five architectures from the Convolutional Neural Network (CNN) method: LeNet, AlexNet, VGG16, Inception V3, and ResNet50. The total dataset used is 500 image data, with the distribution of training data by 80% and test data by 20%. The segmentation process is performed using the Grabcut algorithm by separating the foreground and background. To build a model for CNN architecture using Google Colab and Google Drive storage. The results of the tests carried out on five classes of Aglaonema images obtained the best accuracy, precision, and recall results on the Inception V3 architecture with values of 92.8%, 93%, and 92.8%. The CNN architecture has the highest level of accuracy in classifying aglaonema plant types based on images. This study seeks to close research gaps, contribute to the field of research, and serve as a platform for primary prevention research.

Identification of Paddy Rice Diseases Using Deep Convolutional Neural Networks

In modern digital agricultural applications, automatic identification and diagnosis of plant diseases using artificial intelligence is becoming popular and widespread. Deep learning is a promising tool in pattern recognition and machine learning and it can be used to identify and classify diseases in paddy rice. In this study, 2 different paddy rice diseases, including rice blast and brown spot, were investigated in the district of İpsala in the province of Edirne between the 2020 and 2021 production seasons by collecting 1569 images. These diseases are very common and important in Edirne province and surrounding rice production areas. Therefore, practical methods are needed to identify and classify these two diseases. A Convolutional Neural Network (CNN) model was created by applying pre-processing techniques such as rescaling, rotation, and data augmentation to the paddy rice disease images. The classification model was created in Google Colab, which is a web-based Python editor using Tensorflow and Keras libraries. The CNN model was able to classify rice blast and brown spot diseases with high accuracy of 91.70%.

A Model of a System of Simultaneous Equations with a Lag Effect for Estimating the Quality of an Advertising Campaign

This article describes the creation of a more generalized system of simultaneous equations for forecasting the level of sales depending on advertising campaigns on different channels and other factors. RStudio (R programming language) and Google Colab (Python programming language) environments describe the creation of a model based on real data of a product. The Hausman specification test was applied to determine the model estimation method. As a conclusion, the indicators of advertising campaigns turned out to be endogenous variables, which indicates the importance of using the 2MNK method. It was found that the volume of advertising is the cause of the volume of sales according to Granger, which cannot be said about the reverse assumption - the causality of the volume of advertising from sales according to Granger. The ”depth” of lags is also determined, namely, one lag for both advertising channels. The dependence of sales volumes on various factors, including product distribution, the price index, the influence of advertising and its lags, and the influence of competitors’ advertising activities, was evaluated. The coefficients of the resulting more generalized system of simultaneous equations were estimated using the two-step least squares method. All statistical indicators testify to the adequacy of the model. Performance indicators (ROI - return on investment) of advertising campaigns showed that advertising both on television and on the Internet is profitable for the company’s product in question. The relevance of this article lies in the creation of a more general system of simultaneous equations with the inclusion of a product sales forecast model taking into account the influence of advertising.

Interactive Python Notebooks for Physical Chemistry

Chemistry simulations using interactive graphic user interfaces (GUIs) represent uniquely effective and safe tools to support multidimensional learning. Computer literacy and coding skills have become increasingly important in the chemical sciences. In response to both of these facts, a series of Jupyter notebooks hosted on Google Colaboratory were developed for undergraduate students enrolled in physical chemistry. These modules were developed for use during the COVID-19 pandemic when Millsaps College courses were virtual and only virtual or online laboratories could be used. These interactive exercises employ the Python programming language to explore a variety of chemical problems related to kinetics, the Maxwell–Boltzmann distribution, numerical versus analytical solutions, and real-world application of concepts. All of the modules are available for download from GitHub (https://github.com/Abravene/Python-Notebooks-for-Physical-Chemistry). Accessibility was prioritized, and students were assumed to have no prior programming experience; the notebooks are cost-free and browser-based. Students were guided to use widgets to build interactive GUIs that provide dynamic representations, immediate access to multiple investigations, and interaction with key variables. To evaluate the perceived effectiveness of this introduction to Python programming, participants were surveyed at the beginning and end of the course to gauge their interest in pursuing programming and data analysis skills and how they viewed the importance of programming and data analysis for their future careers. Student reactions were generally positive and showed increased interest in programming and its importance in their futures, so these notebooks will be incorporated into the in-person laboratory in the future.

Android Malware Classification Using LSTM Model

From last two decades, smartphone use is essentially widespread around the world, and Android is the most popular open-source operating system, with the largest market share and active user population of any open-source operating system. This has resulted in malicious actors turning their attention toward the Android operating system to exploit user reliance and vulnerabilities that exist inside the system. Hackers can take advantage of consumers' sensitive data to engage in advertising, extortion, and theft. Most of the existing anti-malware software’s cannot be able to detect all the malwares because of the intelligent malwares. In this paper we use the deep learning based Long short-term memory (LSTM) network for android malware classification. The model is effective in classification of intelligent malwares. The proposed model is implemented using google colab. The model is archiving more than the 98% accuracy in classification of android malwares.

IoT based optical coherence tomography retinal images classification using OCT Deep Net2

Machine learning algorithms gains prominence in health care sectors for disease diagnosis, classification and prediction. Deep learning architecture gains prominence in real time applications. This research work proposes a novel deep learning architecture, OCT Deep Net2 for the classification of optical coherence tomography images. Four class disease classification was performed in this research work and the proposed deep learning framework OCT Deep Net2 is an extension of OCT Deep Net1 comprising of 30 layers. The OCT Deep Net2 comprises of 50 layers and is termed as dense architecture, comprises of three recurrent modules. The performance validation reveals the efficiency of the OCT Deep Net2 architecture in terms of the performance metrics. Robust results were produced for batch size of 32 and 100 epochs with an accuracy of 98%. An IoT based system was implemented using Raspberry Pi B+ processor, the OCT Deep Net 2 algorithm was written in Python and executed on Google Colab.

A deep learning-based approach to facilitate the as-built state recognition of indoor construction works

PurposeRecognising the as-built state of construction elements is crucial for construction progress monitoring. Construction scholars have used computer vision-based algorithms to automate this process. Robust object recognition from indoor site images has been inhibited by technical challenges related to indoor objects, lighting conditions and camera positioning. Compared with traditional machine learning algorithms, one-stage detector deep learning (DL) algorithms can prioritise the inference speed, enable real-time accurate object detection and classification. This study aims to present a DL-based approach to facilitate the as-built state recognition of indoor construction works.Design/methodology/approachThe one-stage DL-based approach was built upon YOLO version 4 (YOLOv4) algorithm using transfer learning with few hyperparameters customised and trained in the Google Colab virtual machine. The process of framing, insulation and drywall installation of indoor partitions was selected as the as-built scenario. For training, images were captured from two indoor sites with publicly available online images.FindingsThe DL model reported a best-trained weight with a mean average precision of 92% and an average loss of 0.83. Compared to previous studies, the automation level of this study is high due to the use of fixed time-lapse cameras for data collection and zero manual intervention from the pre-processing algorithms to enhance visual quality of indoor images.Originality/valueThis study extends the application of DL models for recognising as-built state of indoor construction works upon providing training images. Presenting a workflow on training DL models in a virtual machine platform by reducing the computational complexities associated with DL models is also materialised.

ILEE: Algorithms and toolbox for unguided and accurate quantitative analysis of cytoskeletal images.

The eukaryotic cytoskeleton plays essential roles in cell signaling and trafficking, broadly associated with immunity and diseases in humans and plants. To date, most studies describing cytoskeleton dynamics and function rely on qualitative/quantitative analyses of cytoskeletal images. While state-of-the-art, these approaches face general challenges: the diversity among filaments causes considerable inaccuracy, and the widely adopted image projection leads to bias and information loss. To solve these issues, we developed the Implicit Laplacian of Enhanced Edge (ILEE), an unguided, high-performance approach for 2D/3D-based quantification of cytoskeletal status and organization. Using ILEE, we constructed a Python library to enable automated cytoskeletal image analysis, providing biologically interpretable indices measuring the density, bundling, segmentation, branching, and directionality of the cytoskeleton. Our data demonstrated that ILEE resolves the defects of traditional approaches, enables the detection of novel cytoskeletal features, and yields data with superior accuracy, stability, and robustness. The ILEE toolbox is available for public use through PyPI and Google Colab.

Capturing Provenance from Deep Learning Applications Using Keras-Prov and Colab: a Practical Approach

Due to the exploratory nature of DNNs, DL specialists often need to modify the input dataset, change a filter when preprocessing input data, or fine-tune the models’ hyperparameters, while analyzing the evolution of the training. However, the specialist may lose track of what hyperparameter configurations have been used and tuned if these data are not properly registered. Thus, these configurations must be tracked and made available for the user’s analysis. One way of doing this is to use provenance data derivation traces to help the hyperparameter’s fine-tuning by providing a global data picture with clear dependencies. Current provenance solutions present provenance data disconnected from W3C PROV recommendation, which is difficult to reproduce and compare to other provenance data. To help with these challenges, we present Keras-Prov, an extension to the Keras deep learning library to collect provenance data compliant with PROV. To show the flexibility of Keras-Prov, we extend a previous Keras-Prov demonstration paper with larger experiments using GPUs with the help of Google Colab. Despite the challenges of running a DBMS with virtual environments, DL analysis with provenance has added trust and persistence in databases and PROV serializations. Experiments show Keras-Prov data analysis, during training execution, to support hyperparameter fine-tuning decisions, favoring the comparison, and reproducibility of such DL experiments. Keras-Prov is open source and can be downloaded from https://github.com/dbpina/keras-prov.

Tools of the Google Colaboratory cloud environment for programming training and the preparation for the unified state exam in informatics

One of the cloud technologies that can be used to teach programming based on the Python programming language is the Google Colaboratory cloud service. This service can be used by teachers both when teaching certain themes of the section "Algorithms and programming" as part of a school informatics course, and when preparing students for the tasks in programming of the Unified State Exam in informatics. The article discusses the possibilities of the Google Colaboratory cloud service for creating interactive notebooks using the example of an interactive notebook and practical work on the theme "A precondition loop", as well as interactive notebooks "Task No. 24 of the Unified State Exam in informatics" for students and teachers. The applicability of these notebooks for programming training based on the Python language as part of a school informatics course and for preparing students for tasks in programming of the Unified State Exam in informatics is substantiated.

Automatic Mapping of Burned Areas Using Landsat 8 Time-Series Images in Google Earth Engine: A Case Study from Iran

Due to the natural conditions and inappropriate management responses, large part of plains and forests in Iran have been burned in recent years. Given the increasing availability of open-access satellite images and open-source software packages, we developed a fast and cost-effective remote sensing methodology for characterizing burned areas for the entire country of Iran. We mapped the fire-affected areas using a post-classification supervised method and Landsat 8 time-series images. To this end, the Google Earth Engine (GEE) and Google Colab computing services were used to facilitate the downloading and processing of images as well as allowing for effective implementation of the algorithms. In total, 13 spectral indices were calculated using Landsat 8 images and were added to the nine original bands of Landsat 8. The training polygons of the burned and unburned areas were accurately distinguished based on the information acquired from the Iranian Space Agency (ISA), Sentinel-2 images, and Fire Information for Resource Management System (FIRMS) products. A combination of Genetic Algorithm (GA) and Neural Network (NN) approaches was then implemented to specify 19 optimal features out of the 22 bands. The 19 optimal bands were subsequently applied to two classifiers of NN and Random Forest (RF) in the timespans of 1 January 2019 to 30 December 2020 and of 1 January 2021 to 30 September 2021. The overall classification accuracies of 94% and 96% were obtained for these two classifiers, respectively. The omission and commission errors of both classifiers were also less than 10%, indicating the promising capability of the proposed methodology in detecting the burned areas. To detect the burned areas caused by the wildfire in 2021, the image differencing method was used as well. The resultant models were finally compared to the MODIS fire products over 10 sampled polygons of the burned areas. Overall, the models had a high accuracy in detecting the burned areas in terms of shape and perimeter, which can be further implicated for potential prevention strategies of endangered biodiversity.

Deteksi Kartu Tanda Mahasiswa UIN Pada Gambar dengan Feature Matching

UIN Walisongo's student identity card (KTM) does not have much function other than just for student identification. Even if the function is increased, it can be used for absenteeism at lectures, borrowing books, or double as an automated teller card (ATM). Doing absences using KTM requires a feature matching method for matching the intended KTM image with the KTM that is searched for in the student database. The feature matching process is based on feature detection in images using various methods such as ORB and Scale Invariant Feature Transform (SIFT). We can perform the feature matching method using the Brute-Force method and the Fast Library Approximated Nearest Neighbor (FLANN) on Google Colab with Python. The results of feature matching on the FLANN method are more than the Brute-Force method. The validation of the two image feature matching was carried out using the Root Mean Square Error (RMSE) method, resulting in an average value of 10.424. The purpose of this article is to detect student ID cards with matching features in the image. The FLANN and Brute-Force feature matching methods can be used to detect KTM UIN in images.
 Keywords: feature matching, SIFT, FLANN, Brute-Force, RMSE