Google Colab Research Articles

Convolutional neural networks in the qualitative improvement of sweet potato roots

The objective was to verify whether convolutional neural networks can help sweet potato phenotyping for qualitative traits. We evaluated 16 families of sweet potato half-sibs in a randomized block design with four replications. We obtained the images at the plant level and used the ExpImage package of the R software to reduce the resolution and individualize one root per image. We grouped them according to their classifications regarding shape, peel color, and damage caused by insects. 600 roots of each class were destined for training the networks, while the rest was used to verify the quality of the fit. We used the python language on the Google Colab platform and the Keras library, considering the VGG-16, Inception-v3, ResNet-50, InceptionResNetV2, and EfficientNetB3 architectures. The InceptionResNetV2 architecture stood out with high accuracy in classifying individuals according to shape, insect damage, and peel color. Image analysis associated with deep learning may help develop applications used by rural producers and improve sweet potatoes, reducing subjectivity, labor, time, and financial resources in phenotyping.

Deteksi Atap Bangunan Berbasis Citra Udara Menggunakan Google Colab dan Algoritma Deep Learning YOLOv7

This article presents a deep learning-based approach for roof building detection in aerial imagery using the Yolov7 algorithm and Google Colab. We trained the Yolov7 model on a dataset of annotated aerial images and tested its performance on a testing set of 200 images. Our experimental results show that the Yolov7 model achieved an F1 score of 80% for roof building detection on the testing set, indicating that our approach can accurately detect roof buildings in aerial imagery. This approach can be useful in various applications, such as urban planning, disaster management, and infrastructure development. The availability of aerial imagery and the ability to use deep learning algorithms make it possible to automate the process of roof building detection, which can save time and resources. This research can be extended by investigating the use of other deep learning algorithms or combining multiple algorithms to improve the accuracy of roof building detection in aerial imagery.

Analisis Kepuasan Learning Management System Universitas XYZ Menggunakan Metode System Usability Scale dan K-Means

The importance of knowing the results of using the LMS (Learning Management System) learning application to determine the overall use value of users each semester. User perceptions were obtained using the SUS (system usability scale) method using a questionnaire that adopted 10 questions distributed to 118 users (students) of the XYZ University Informatics Engineering study program who had used the LMS after 1 semester. The purpose of this study is to determine user perceptions by clustering user satisfaction which has been carried out for 1 semester. Grouping perceptions using the K-Means method with variables (columns) that seem to have the greatest influence on other variables. Other tools use Google Colab in the Python programming language. The number of variables is 10 variables adopted from the questions in the System Usability Scale method. The results of this study provide a total of 3 (three) clusters which will then become the basis for scoring the criteria for the SUS method. The criteria for using the LMS system with cluster 2 have an excellent rating (SUS score of 72.04) and the number of perceptions is 49 people from 118 students. Overall, LMS users provide good value for several modules in the LMS, but the third cluster with the highest number gives the best results from the other clusters.

Pharmacoinformatic Investigation of Silymarin as a Potential Inhibitor against Nemopilema nomurai Jellyfish Metalloproteinase Toxin-like Protein.

Jellyfish stings pose a major threat to swimmers and fishermen worldwide. These creatures have explosive cells containing one large secretory organelle called a nematocyst in their tentacles, which contains venom used to immobilize prey. Nemopilema nomurai, a venomous jellyfish belonging to the phylum Cnidaria, produces venom (NnV) comprising various toxins known for their lethal effects on many organisms. Of these toxins, metalloproteinases (which belong to the toxic protease family) play a significant role in local symptoms such as dermatitis and anaphylaxis, as well as systemic reactions such as blood coagulation, disseminated intravascular coagulation, tissue injury, and hemorrhage. Hence, a potential metalloproteinase inhibitor (MPI) could be a promising candidate for reducing the effects of venom toxicity. For this study, we retrieved the Nemopilema nomurai venom metalloproteinase sequence (NnV-MPs) from transcriptome data and modeled its three-dimensional structure using AlphaFold2 in a Google Colab notebook. We employed a pharmacoinformatics approach to screen 39 flavonoids and identify the most potent inhibitor against NnV-MP. Previous studies have demonstrated the efficacy of flavonoids against other animal venoms. Based on our analysis, Silymarin emerged as the top inhibitor through ADMET, docking, and molecular dynamics analyses. In silico simulations provide detailed information on the toxin and ligand binding affinity. Our results demonstrate that Silymarin's strong inhibitory effect on NnV-MP is driven by hydrophobic affinity and optimal hydrogen bonding. These findings suggest that Silymarin could serve as an effective inhibitor of NnV-MP, potentially reducing the toxicity associated with jellyfish envenomation.

The Analytical Flory Random Coil Is a Simple-to-Use Reference Model for Unfolded and Disordered Proteins.

Denatured, unfolded, and intrinsically disordered proteins (collectively referred to here as unfolded proteins) can be described using analytical polymer models. These models capture various polymeric properties and can be fit to simulation results or experimental data. However, the model parameters commonly require users' decisions, making them useful for data interpretation but less clearly applicable as stand-alone reference models. Here we use all-atom simulations of polypeptides in conjunction with polymer scaling theory to parameterize an analytical model of unfolded polypeptides that behave as ideal chains (ν = 0.50). The model, which we call the analytical Flory random coil (AFRC), requires only the amino acid sequence as input and provides direct access to probability distributions of global and local conformational order parameters. The model defines a specific reference state to which experimental and computational results can be compared and normalized. As a proof-of-concept, we use the AFRC to identify sequence-specific intramolecular interactions in simulations of disordered proteins. We also use the AFRC to contextualize a curated set of 145 different radii of gyration obtained from previously published small-angle X-ray scattering experiments of disordered proteins. The AFRC is implemented as a stand-alone software package and is also available via a Google Colab notebook. In summary, the AFRC provides a simple-to-use reference polymer model that can guide intuition and aid in interpreting experimental or simulation results.

Experimental Validation Through a Parallel Computation Algorithm for Evaluation Uncertainty of the Mathematical Model of Direct Expansion Solar Assisted Heat Pump

This paper presents the development of a mathematical model for a direct-expansion solar-assisted heat pump (DX-SAHP) operating in steady-state. The mathematical model was implemented using the scientific software EES and using a code written in Python. It was utilized a lumped parameter model for the heat exchangers and a semi-empirical model for the compressor. The mathematical model was validated using experimental data of a DX-SAHP running with R134a. Two hundred simulations were made combining different correlations for estimating the convective heat transfer coefficient in the evaporator/collector. The Mean Absolute Deviation (MAD) and the Mean Deviation (MD) between the theoretical and experimental values for the COP were 2.6±1.8 % and 0.9±1.8 %, respectively. The MAD and MD between the discharge temperature were 1.56±0.16 % and -1.45±0.16 %. The mean difference between the results using EES and Python were 1.4 %. The use of Python with parallel computing for uncertainty analyses, reduced the simulation time in 88 % if compared with EES. The model in Python is available as open-source through the platform Google Colaboratory.

Deep learning models used for accuracy and loss detection on twitter dataset by setting tuning parameters

Hate speech on social media post is running now a days. Social media like YouTube, Twitter, and Facebook etc. are responsible for hated speech. Hated speech spreads through digital media, causing individuals to get confused and adopt prejudiced viewpoints. To limit the negative effects of disinformation on the digital platform, it is critical to detect it. Now a days, lots of digital platforms are available. Hate speech detection in dataset is very difficult. As a result, the Twitter dataset is of the size of 25296 is presented in this work. Many deep learning techniques are applied on Twitter dataset. The Google Colab tool is used to scrape dataset material. Different deep learning approaches are utilized to boost the accuracy of the hated speech dataset. For training and validation accuracy and loss some models are used on Twitter dataset like Bi-directional Long Short Term Memory with Glove, Bi-LSTM, and Embedding from Language Model (Elmo) with deep learning, Convolutional Neural Network (CNN), Long Short Term Memory with Glove and LSTM. The performance of the proposed tweet dataset is evaluated using a variety of deep learning classifiers on text dataset. The planned deep learning techniques produced good results on tweet dataset. LSTM with Glove gave the highest accuracy 0.89 and minimum loss 0.19 on tweet dataset. So when compare our model on same dataset that was used earlier then we get highest accuracy and minimum loss.

Vision-Based Soft Mobile Robot Inspired by Silkworm Body and Movement Behavior

Designing an inexpensive, low-noise, safe for individual, mobile robot with an efficient vision system represents a challenge. This paper proposes a soft mobile robot inspired by the silkworm body structure and moving behavior. Two identical pneumatic artificial muscles (PAM) have been used to design the body of the robot by sewing the PAMs longitudinally. The proposed robot moves forward, left, and right in steps depending on the relative contraction ratio of the actuators. The connection between the two artificial muscles gives the steering performance at different air pressures of each PAM. A camera (eye) integrated into the proposed soft robot helps it to control its motion and direction. The silkworm soft robot detects a specific object and tracks it continuously. The proposed vision system is used to help with automatic tracking based on deep learning platforms with real-time live IR camera. The object detection platform, named, YOLOv3 is used effectively to solve the challenge of detecting high-speed tiny objects like Tennis balls. The model is trained with a dataset consisting of images of Tennis balls. The work is simulated with Google Colab and then tested in real-time on an embedded device mated with a fast GPU called Jetson Nano development kit. The presented object follower robot is cheap, fast-tracking, and friendly to the environment. The system reaches a 99% accuracy rate during training and testing. Validation results are obtained and recorded to prove the effectiveness of this novel silkworm soft robot. The research contribution is designing and implementing a soft mobile robot with an effective vision system.

Kıyma Kokuşma Analizi İçin Düşük Maliyetli Elektronik Burun Geliştirilmesi

A low-cost, easy-to-use e-nose is developed to detect the spoilage of ground meat. E-nose consists of hardware, software and data processing components. The main elements of hardware component are gas sensors sensitive to hydrogen sulfide (H2S) and ammonia (NH3). Using MIT App Inventor 2 an Android application is developed to run the hardware component, retrieve the data, pre-process and send it to Google Sheets. Classification model is developed, and data management is carried out in Google Colab and Google Script. Logistic regression method is used to develop classification models from the collected signals. The model classified the samples as "spoiled" and "fresh" based on the gas concentrations. The Nessler solution is used to determine the actual spoilage state. Ground beef samples stored in the refrigerator and at room temperature are used to obtain spoiled and fresh samples to develop a logistic regression model. A total of 36 samples are used to develop model. Another set of 24 samples is used to test model and prototype device performance. It is observed that all samples used in the testing phase were classified correctly. The cost of the system has been determined as approximately $100 considering January 2021 exchange rates.

Mathematical Model of Two-Dimensional Disease Progression and Categorization of Underlying Vector Fields with Machine Learning

A mathematical model was created for simulating disease progression in two-dimension, by specifying a vector field. It was equivalent to observing how a lesion pattern on the skin, could change its shape over time. This model had a parameter that might control the resolution and signal of the pattern. By specifying two different vector fields, we established two sets of subtly different image sets. And, three different supervised machine learning algorithms such as nearest neighbor, decision tree, and support vector machine were used for binary classification. This study was done with Python's numerical (numpy), plotting (matplotlib), and machine learning (sklearn) libraries on the Google Colab platform. All machine learning algorithms were able to distinguish subtle differences produced by different vector fields. Furthermore, the performance of the algorithms were improved by concatenating the beginning and end stages of the pattern and helping the algorithms to pick up the temporal changes. These results demonstrated how AI and machine learning could be adopted in medicine for accurately diagnosing underlying diseases from images.

Performance evaluation of YOLO v5 model for automatic crop and weed classification on UAV images

For sustainability and efficiency in maintaining high crop yield and less chemically polluted agricultural lands, precise weed mapping is essential for the total implementation of site-specific weed management which currently stands as a major challenge in present day agriculture. In this research, the robustness of the training epochs of You Only Look Once (YOLO) v5s, a Convolutional Neural Network (CNN) model was evaluated for the development of an automatic crop and weeds classification using UAV images. The images were annotated using a bounding box and they were trained on google colaboratory over 100, 300, 500, 600, 700 and 1000 epochs. The model detected and categorized five different classes which are sugarcane (Saccharum officinarum), banana trees (Musa), spinach (Spinacia oleracea), pepper (Capsicum), and weeds. To find the optimal performance on the test set, the model was trained across several epochs, and training was stopped when the test performance (classification accuracy, precision, and recall) began to drop. The obtained result shows that the performance of the classifier improved significantly as the range of training epochs tends to rise from 100 through to 600 epochs. Meanwhile, a slight decline was observed as the number of epoch was increased to 700 when the classification accuracy, the precision of weed and recall of 65, 43 and 43%, respectively, was recorded as against 67, 78 and 34% that was obtained as the classification accuracy, weed precision and recall, respectively, at 600 epochs. This decline continued even when the epoch was increased to 1000 where classification accuracy, weed precision and recall of 65%, 45% and 40%, respectively was obtained. The results showed that the training epoch of YOLOv5s significantly affects the model's robustness in automatic crop and weep classification and identified 600 as the epoch for optimal performance.

SWOT Analysis of Purse Seine Fishery Business

Objectives - Identify the internal and external environment of the Purse Seine Catching Business which includes micro and macro environments and formulate alternative business development strategies for the purse seine catching business. Design/Methodology/Approach – Sampling using a questionnaire with 5 people who are experts or understand including business owners, Fishing Masters, Engineers, Marketing Department, and Boat Fishermen. Testing data analysis using the R language through Google Colab. Findings - These businesses are strong and likely, the recommendations given are Progressive, meaning that businesses are run in prime and steady condition so that it is possible to continue to expand, increase growth and achieve maximum progress. Originality/value – The added value of this research contributes to business development explaining what strategies a company should take to maximize opportunities and minimize threats

Comparative Analysis of Machine Learning Performance in House Price Prediction for Ames Iowa

The price of housing is closely related to the quality of people's lives, which is reflected in various aspects (e.g., finding a place to live and managing their finances). On this basis, being able to predict house prices would be a great convenience, and this study examines the accuracy of multiple mechanistic learning regression models in predicting future house price changes. To be specific, this paper downloaded the dataset from Kaggle which includes the training and testing csv files. and ran many different regression techniques including lasso and more models on google Colab. Furthermore, a comparison between the results of different methods is carried out by estimating the mean square error. Then, this research combined some of the best yielded results of the model to combine into a final model and got a result with a high score. According to the analysis and evaluations, the study concluded that the hybrid model resulted in a more accurate prediction than the single model. These results shed light on guiding further exploration of real estate prediction.

Real-time multiple face mask and fever detection using YOLOv3 and TensorFlow lite platforms

COVID-19 has caused disruptions to many aspects of everyday life. To reduce the impact of this pandemic, its spreading must be controlled via face mask wearing. Manually mask-checking for everybody is embarrassing and uncontrollable. Hence, the proposed technique is used to help for automatic mask-checking based on deep learning platforms with real-time surveillance live infra-red (IR) camera. In this paper, two recent object detection platforms, named, you only look once version 3 (YOLOv3) and TensorFlow lite are adopted to accomplish this task. The two models are trained with a dataset consisting of images of persons with/without masks. This work is simulated with Google Colab then tested in real-time on an embedded device mated with fast GPU called Raspberry Pi 4 model B, 8 GB RAM. A comparison is made between the two models to verify their performance in relation to their precision rate and processing time. The work of this paper is also succeeded to realize multiple face masks real-time detection up to 10 facemasks in a single scene with high inference speed. Temperature is also measured using IR touchless sensor for each person with sound alarming to alert fever. The presented detector is cheap, light, small, and fast, with 99% accuracy rate during training and testing.

A Simple Fuzzy Decision Support Model for Determining the Indonesian Green Cities

Green city is every bodies’ expectation. Every country highly competes to increase the number of green city. Thus several types of measurement standard for city green level appear and being practically implemented. The study’s objective is to develop a decision support model (DSM) for determining the Indonesian green cities based on the main method fuzzy logic. At this juncture, the Indonesian green city determined by green score of each city in Indonesia. The main method fuzzy logic interconnected to statistics and mathematics approaches. Slope calculation that combined with fuzzy logic conception was operated to treat three of five parameters believed in the model. The model design was delivered by using the object oriented approach and the model itself constructed via programming language Python in Google Colab platform. Finally, the constructed model is able to exhibit a hundred Indonesian cities’ green scores.

Secure Identification at Your Fingertips: Building a Face Recognition System with Google Colab

Abstract: The term "security" refers to the measures taken to ensure protection against harm. A security system is the result of measures taken to secure something through a system. There are various types of security systems available today, and ongoing research is being conducted in this field. The purpose of this research paper is to develop a face recognition system to enhance security, with the aim of making it more reliable and secure than existing systems. This will be achieved through the use of Google Colaboratory (Colab) for face recognition, which will enable the system to accurately identify individuals based on their facial features. The system will automatically distinguish between real and fake samples to prevent unauthorized access attempts. If a fake access attempt is detected, the system will display a message indicating that the user is unauthorized. The proposed method aims to improve accuracy, efficiency, clarity, and security, ultimately enhancing the overall security of the system.

Pengembangan Sistem Pendeteksi Jenis Sayuran dengan Metode CNN Berbasis Android

Vegetables are foodstuffs of plant origin that can be consumed fresh and have various health benefits. However, not a few people do not know the types of vegetables and will find it difficult to find the vegetables they want. This research aims to make it easier for people to find vegetables by classifying them. Researchers developed a model using the Convolutional Neural Network (CNN) method with a total of 15 datasets with a total of 3000 image data. Researchers conducted training datasets with 3 types of epochs, including 20 epochs, 50 epochs and 100 epochs. The training produces accuracy and training loss, with the highest accuracy belonging to Epoch 50 and Epoch 100 and the lowest level of training loss is owned by Epoch 100 with a total of 0.609. However, after the model was deployed, the accuracy results obtained were not as high as the tests conducted on Google Colab. Tests were carried out on several objects, including carrots with an accuracy of 69%, cabbage with an accuracy of 53%, and papaya with an accuracy of 82%. The difference in accuracy results may be caused by objects that are less identical to the datasets or can also be caused by imperfect models. Even so, this application can already be used to classify types of vegetables.

읽기 동기 수준에 따른 고등학생의 추론적 이해의 차이

Objectives The purpose of this study is to analyze the difference in high school students' inferential comprehension according to the level of reading motivation. Methods Reading motivation and inferential comprehension of 517 high school students were measured using reading motivation and inferential comprehension tests. Based on the response results, descriptive statistics and correlations were confirmed using Google's colab, and the effect size was calculated with Cohen's d. Results First, the correlation coefficient between reading motivation and inferential comprehension was .308, and the reading efficacy factor showed the highest correlation with inferential comprehension. A difference of about 0.51 was confirmed when calculating the effect size for inferential understanding between a group with high reading motivation and a group with low reading motivation. Second, the correlation between reading motivation and inferential comprehension was high in both the high and low reading motivation groups. Third, when calculated by the effect size, a difference of about 0.19 was confirmed in the inferential comprehension by gender of the group with high motivation to read, but a difference of about .01 was confirmed in the inferential comprehension by gender in the group with low motivation to read. Conclusions First, it is necessary to consider reading motivation when teaching reasoning comprehension, and it is necessary to check reading motivation in advance and design teaching and learning considering it when teaching reasoning comprehension targeting students with low reading motivation. Second, when teaching literature subjects that mainly deal with narrative texts, it is necessary to check students' reading motivation and consider it, and the use of narrative texts can be effective when teaching inferential understanding. Third, if female students are expected to have a higher level of inferential understanding than male students, and instruction in reading may not achieve the intended educational effect.

Breast cancer detection from histopathological image dataset using hybrid convolution neural network

Cancer of the breast is a deadly disease that can take a person’s life in many different ways. Predicting breast cancer at an early stage is crucial in the fight to end the disease. The usage of deep learning and blockchain technology has been implemented with the intention of integrating optimal prediction with clinical diagnostics and protecting private health information. Patients’ medical records are encrypted and stored in the blockchain for maximum safety. As a result, a large portion of time and energy is spent on feature engineering, a tedious process that requires prior expert domain knowledge of the data to develop effective features, and is crucial to the success of most conventional classification systems. Deep learning, on the other hand, can arrange the discriminative information in the data without the need for a domain expert to develop feature extractors. The research community and industry have paid attention to deep, feedforward networks like convolutional neural networks (CNNs) because of their empirical results in areas including speech recognition, signal processing, object recognition, natural language processing, and transfer learning. For the best breast cancer prediction, a new method called a “Hybrid CNN” combining the Sine Cosine Algorithm (SCA) with a transfer learning algorithm has been presented. Mini-batch size and drop-out rate are just two of the factors that the SCA algorithm may fine-tune. To stop the model from overfitting, we employ a transfer learning strategy. The hyperparameters found using sine cosine algorithm is used in Visual geometry Group (VGG 16) architecture. ImageNet is used to pretrain the network and last three convolutional layers are trained using transfer learning. The integration of sine cosine algorithm and transfer learning together increases the accuracy thereby preventing the model from overfitting. The experimentation is performed in Google Colab and the proposed Hybrid CNN is compared with existing methodologies such as K-NN, SVM. Also, the proposed Hybrid CNN is compared with CNN without transfer learning and CNN without SCA. The metrics taken into account for comparison are accuracy, sensitivity, specificity, F-score. The proposed Hybrid CNN achieves 96.9% accuracy that shows the effectiveness of the integration of SCA and transfer learning.

A Study on an Effective Teaching of AI using Google Colab-Based DCGAN Deep Learning Model Building for Music Data Analysis and Genre Classification

This paper deals with an effective teaching method of deep learning using theory and Python in the University. Currently, AI and related technology penetrate into all areas such as manufacturing, fashion, design, medical, novel, agriculture, as well as picture and engineering areas. These AI technologies are strongly connected with the education of universities and K-12. There are two categories of AI-related education. The first one is AI-supported education; another thing is education (teaching and learning) to understand AI. In any case, AI and its application method should be taught with theory and performed with S/W. This paper provides a method on how teachers of universities can teach deep learning well with S/W (Python) matching theory. To present the characteristics of deep learning, this paper uses DCGAN and suggests a teaching method with Google Colab easily. This paper analyzes the dataset with visuals and classifies genres to show characteristics between music and the function of deep learning for students' understanding using DCGAN and the music dataset. The results classify music genres by deep learning well.