Logo Freelancer
Se connecter S'inscrire
Publier un projet
EXPLORER
.NET
C# Programming
PHP
Python
SQL
Photo de couverture du profil
Vous suivez désormais
Erreur de suivi de l'utilisateur.
Cet utilisateur n'utilise pas les utilisateurs à le suivre.
Vous suivez déjà cet utilisateur.
Votre plan d'adhésion ne permet que 0 suivis. Améliorez ici.
Ne suit désormais plus
Erreur lors de l'arrêt du suivi de l'utilisateur.
Vous avez désormais recommandé
Erreur lors de la recommendation de l'utilisateur.
Une erreur a eu lieu. Veuillez rafraîchir la page et réessayer.
E-mail désormais vérifié.
Avatar de l'utilisateur

Farbod M.

@FarWebDigital

0,0
0
0,0
0,0
0%

AI Developer & Web Developer

$20 USD / Heure
・
Drapeau de
Canada (11:51 AM)
・
Inscrit depuis le mai 16, 2021
$20 USD / Heure
・
Hello. I am Farbod. 5+ years of experience in (Machine Learning, Generative models, NLP, Big Data, API, and Back-End Development). Built end-to-end machine learning systems. ✅ Successfully Completed Projects: 1. Image Classification, Image segmentation, Image retrieval 2. Exploratory analysis and data visualization 3. Sentiment Analysis 4. Medical Image analysis 5. Question Answering System 6. Topic Modeling 7. Extract product specifications from product descriptions using GPT-3 8. GPT-3 Prompt design, GPT-3 prompt engineering, Fine-tuning GPT-3
Vérifications
Freelancer PHP Developers Canada
Voir tous les PHP Developers en Canada

Portfolio

Portfolio

11566667
Project description Deep Learning for Automatic Colorization of Grayscale Aerial Photographs using WGAN & Pix2Pix https://github.com/herodesigngit/Deep-Learning-for-Automatic-Colorization-of-Legacy-Grayscale-Aerial-Photographs-using-WGAN-Pix2Pix
Deep-Learning-for-Automatic-Colorization
11566663
Radar-based drone, car, people classification using transfer learning. https://github.com/herodesigngit/Radar-based-drone-classification
Radar-based drone, car, people classification using transfer
11566660
RFM, also known as RFM analysis, is a type of customer segmentation and behavioral targeting used to help businesses rank and segment customers based on the recency, frequency, and monetary value of a transaction. The shopping cart information of the customers of a chain store is available in an Excel file for a period of 1 year. In this project, analysis will be done on this information. This information includes all the variables that are recorded in each customer's purchase invoice, including customer name, product name, customer ID, product ID, product number, product amount, total invoice amount, store branch number, etc. Elbow diagram is used to determine the number of clusters and silhouette index is used to determine the quality of clustering. To determine the quality of the final clusters, ANOVA index is used, and then CLV based on Shannon's RFMC weighting method is calculated for each cluster. Next, CLV based on Shannon's RFMC weighting method is calculated for each cluster.
Customer segmentation using RFMC analysis
11566652
Project description Customer churn prediction for insurance company. This project is for an insurance company with real data and we are looking for a model to predict customer churn and model it. https://github.com/herodesigngit/Churn-Prediction
Customer churn prediction for insurance company
Project description Customer churn prediction for insurance company. This project is for an insurance company with real data and we are looking for a model to predict customer churn and model it. https://github.com/herodesigngit/Churn-Prediction
Customer churn prediction for insurance company
11566630
Role Research scientist and python programmer Project goal Increasing the performance of skin lesion classification system through optimizing hyper-parameters and architecture of deep neural network using metaheuristic optimization algorithms. Solution Three optimization algorithms are employed to find an optimal configuration for the convolutional neural network either in pre-trained models or model that are trained from scratch. Then the deep features extracted from the optimized models were fused together in pairs and used to train a KNN classifier.The effect of applying hyper-parameter optimization is evaluated on ISIC 2017 and ISIC 2018 datasets. The accuracy of the deep neural network produced by our method reaches to 81.6% and F1-score of 80.9% on ISIC 2017 dataset and accuracy of 90.1% and F1-score of 89.8% on ISIC 2018.
Hyper parameter optimization for skin lesion classification
Role Research scientist and python programmer Project goal Increasing the performance of skin lesion classification system through optimizing hyper-parameters and architecture of deep neural network using metaheuristic optimization algorithms. Solution Three optimization algorithms are employed to find an optimal configuration for the convolutional neural network either in pre-trained models or model that are trained from scratch. Then the deep features extracted from the optimized models were fused together in pairs and used to train a KNN classifier.The effect of applying hyper-parameter optimization is evaluated on ISIC 2017 and ISIC 2018 datasets. The accuracy of the deep neural network produced by our method reaches to 81.6% and F1-score of 80.9% on ISIC 2017 dataset and accuracy of 90.1% and F1-score of 89.8% on ISIC 2018.
Hyper parameter optimization for skin lesion classification
11566667
Project description Deep Learning for Automatic Colorization of Grayscale Aerial Photographs using WGAN & Pix2Pix https://github.com/herodesigngit/Deep-Learning-for-Automatic-Colorization-of-Legacy-Grayscale-Aerial-Photographs-using-WGAN-Pix2Pix
Deep-Learning-for-Automatic-Colorization
11566663
Radar-based drone, car, people classification using transfer learning. https://github.com/herodesigngit/Radar-based-drone-classification
Radar-based drone, car, people classification using transfer
11566660
RFM, also known as RFM analysis, is a type of customer segmentation and behavioral targeting used to help businesses rank and segment customers based on the recency, frequency, and monetary value of a transaction. The shopping cart information of the customers of a chain store is available in an Excel file for a period of 1 year. In this project, analysis will be done on this information. This information includes all the variables that are recorded in each customer's purchase invoice, including customer name, product name, customer ID, product ID, product number, product amount, total invoice amount, store branch number, etc. Elbow diagram is used to determine the number of clusters and silhouette index is used to determine the quality of clustering. To determine the quality of the final clusters, ANOVA index is used, and then CLV based on Shannon's RFMC weighting method is calculated for each cluster. Next, CLV based on Shannon's RFMC weighting method is calculated for each cluster.
Customer segmentation using RFMC analysis
11566652
Project description Customer churn prediction for insurance company. This project is for an insurance company with real data and we are looking for a model to predict customer churn and model it. https://github.com/herodesigngit/Churn-Prediction
Customer churn prediction for insurance company
Project description Customer churn prediction for insurance company. This project is for an insurance company with real data and we are looking for a model to predict customer churn and model it. https://github.com/herodesigngit/Churn-Prediction
Customer churn prediction for insurance company
11566630
Role Research scientist and python programmer Project goal Increasing the performance of skin lesion classification system through optimizing hyper-parameters and architecture of deep neural network using metaheuristic optimization algorithms. Solution Three optimization algorithms are employed to find an optimal configuration for the convolutional neural network either in pre-trained models or model that are trained from scratch. Then the deep features extracted from the optimized models were fused together in pairs and used to train a KNN classifier.The effect of applying hyper-parameter optimization is evaluated on ISIC 2017 and ISIC 2018 datasets. The accuracy of the deep neural network produced by our method reaches to 81.6% and F1-score of 80.9% on ISIC 2017 dataset and accuracy of 90.1% and F1-score of 89.8% on ISIC 2018.
Hyper parameter optimization for skin lesion classification
Role Research scientist and python programmer Project goal Increasing the performance of skin lesion classification system through optimizing hyper-parameters and architecture of deep neural network using metaheuristic optimization algorithms. Solution Three optimization algorithms are employed to find an optimal configuration for the convolutional neural network either in pre-trained models or model that are trained from scratch. Then the deep features extracted from the optimized models were fused together in pairs and used to train a KNN classifier.The effect of applying hyper-parameter optimization is evaluated on ISIC 2017 and ISIC 2018 datasets. The accuracy of the deep neural network produced by our method reaches to 81.6% and F1-score of 80.9% on ISIC 2017 dataset and accuracy of 90.1% and F1-score of 89.8% on ISIC 2018.
Hyper parameter optimization for skin lesion classification
Modifications enregistrées
0.0 · 0 Reviews
Commentaires
Aucun commentaire visible ici !
Invitation désormais envoyée !
Merci ! Nous vous avons envoyé un lien par e-mail afin de réclamer votre crédit gratuit.
Une erreur a eu lieu lors de l'envoi de votre e-mail. Veuillez réessayer.
Logo Freelancer France / Français
Aide & Assistance
 Freelance
CatégoriesProjetsConcours FreelancesEntreprise Développement d'IA AdhésionProgramme de Freelance PréféréGestion de projetEmplois locauxPhoto AnywherePrésentationAPI pour les développeurs Obtenir la certification Application Desktop
 À propos
À propos de nousComment cela fonctionneSécuritéInvestisseurPlan du siteHistoiresNouveautésÉquipeRécompensesCommuniqué de presseOffres d'emploi
 Termes
Politique de ConfidentialitéTermes et ConditionsPolitique de CopyrightCode de conduiteFrais et charges
Partenaires
Escrow.comLoadshiftWarrior Forum
 Applications
Logo d'Apple App StoreLogo Google play
Échec de copie dans le presse-papier, veuillez réessayer après avoir ajuster vos permissions.
Copié dans le presse-papier.
Utilisateurs enregistrés Total des travaux publiés
Freelancer ® is a registered Trademark of Freelancer Technology Pty Limited (ACN 142 189 759)
Copyright © 2024 Freelancer Technology Pty Limited (ACN 142 189 759)
Chargement de l'aperçu
Permission donnée pour la géolocalisation.
Votre session de connexion a expiré et vous avez été déconnecté. Veuillez vous connecter à nouveau.