Logo Freelancer Comment ça marche Parcourir les emplois
Se connecter S'inscrire
Publier un projet
EXPLORER
.NET
C# Programming
Cloud Computing
Machine Learning (ML)
Python
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
$20 USD / heure
Drapeau de MALAYSIA
melaka, malaysia
$20 USD / heure
Il est actuellement 11:48 PM ici
Rejoint le octobre 7, 2012
0 Recommandations

Nur Aishah A.

@nuraishahrazak

 5,0 (1 commentaire)
2,7
2,7
100%
100%
$20 USD / heure
Drapeau de MALAYSIA
melaka, malaysia
$20 USD / heure
100 %
Travaux complétés
100 %
Suivant le budget
100 %
Dans les temps
N/A
Taux de réembauche

Software Developer . Machine Learning . AI

Hi there! I would like to help you translate your business needs into solutions. With more than 10 years experience in Software Development and IT Consulting, I have a knack of not just developing Software but also getting very good understanding of business workflows and coming up with end-to-end solutions. I have worked on numerous Desktop Applications (Windows Forms, WPF, WCF web services) and Web Applications (.NET MVC, .NET Core, Backbone, Angular, Azure, GCP, Flask) (hosted both on-prem and on cloud) in my professional life. I am also experienced in Artificial Intelligence - specifically in Computer Vision and Deep Learning. Do drop me a note if you'd like to discuss possible opportunities. I'd be happy to have a quick chat with you.
Freelancer Python Developers Malaysia
Voir tous les Python Developers en Malaysia

Contactez Nur Aishah A. concernant votre travail

Connectez-vous pour discuter de tous les détails via la messagerie.

Portfolio

10345478
These are data analysis and visualizations I developed for my research that looks at one facet of predicting the spread of Covid-19 cases by looking at the relationship between government policies being put in place and the total number of cases by utilizing Machine Learning techniques.
Data Analysis & Visualization for Covid19 Government Measure
These are data analysis and visualizations I developed for my research that looks at one facet of predicting the spread of Covid-19 cases by looking at the relationship between government policies being put in place and the total number of cases by utilizing Machine Learning techniques.
Data Analysis & Visualization for Covid19 Government Measure
These are data analysis and visualizations I developed for my research that looks at one facet of predicting the spread of Covid-19 cases by looking at the relationship between government policies being put in place and the total number of cases by utilizing Machine Learning techniques.
Data Analysis & Visualization for Covid19 Government Measure
These are data analysis and visualizations I developed for my research that looks at one facet of predicting the spread of Covid-19 cases by looking at the relationship between government policies being put in place and the total number of cases by utilizing Machine Learning techniques.
Data Analysis & Visualization for Covid19 Government Measure
10305823
This is a Flask App for analysing Malay Language Morphology. The main advantage of this analyzer is the ability to morph a word into a noun or verb by adding prefixes to the word. Other advantages of the analyzer I created is that it considers edge cases like the word ‘ajar’ that has different prefix to use (‘pe-‘) and one syllable word like ‘bom’ that uses ‘menge-‘ and ‘penge-‘ prefix. This analyzer is also able to change a word into a intransitive verb e.g. ‘berlari’ by prefixing ‘ber-‘ to the input word.
NLP: Malay Language Morphology Analyser
This is a Flask App for analysing Malay Language Morphology. The main advantage of this analyzer is the ability to morph a word into a noun or verb by adding prefixes to the word. Other advantages of the analyzer I created is that it considers edge cases like the word ‘ajar’ that has different prefix to use (‘pe-‘) and one syllable word like ‘bom’ that uses ‘menge-‘ and ‘penge-‘ prefix. This analyzer is also able to change a word into a intransitive verb e.g. ‘berlari’ by prefixing ‘ber-‘ to the input word.
NLP: Malay Language Morphology Analyser
10305803
Ontology Project: Companies' Named Entity Extraction
10305798
The thesis seeks to demonstrate how transfer learning from macro-expression database in a lightweight network can achieve high Micro-Expression Recognition (MER). The proposed MER model is developed from truncated EfficientNet-B0 model that is pre-trained on a macro-expression database before fine-tuned on the CDE dataset. A simple algorithm-tuning that manipulates the loss function is suggested to deal with the imbalanced dataset. The resulting model is considerably small with just 867K parameters. Experimental results using Leave-One-Subject-Out cross validation show substantial performance increase compared to the state-of-the-arts with UF1 score of 0.98 and UAR score of 1.77. The images attached shows the pre-processing steps done to a sample of the dataset.
Thesis: Lightweight Micro-Expression Recognition
The thesis seeks to demonstrate how transfer learning from macro-expression database in a lightweight network can achieve high Micro-Expression Recognition (MER). The proposed MER model is developed from truncated EfficientNet-B0 model that is pre-trained on a macro-expression database before fine-tuned on the CDE dataset. A simple algorithm-tuning that manipulates the loss function is suggested to deal with the imbalanced dataset. The resulting model is considerably small with just 867K parameters. Experimental results using Leave-One-Subject-Out cross validation show substantial performance increase compared to the state-of-the-arts with UF1 score of 0.98 and UAR score of 1.77. The images attached shows the pre-processing steps done to a sample of the dataset.
Thesis: Lightweight Micro-Expression Recognition
The thesis seeks to demonstrate how transfer learning from macro-expression database in a lightweight network can achieve high Micro-Expression Recognition (MER). The proposed MER model is developed from truncated EfficientNet-B0 model that is pre-trained on a macro-expression database before fine-tuned on the CDE dataset. A simple algorithm-tuning that manipulates the loss function is suggested to deal with the imbalanced dataset. The resulting model is considerably small with just 867K parameters. Experimental results using Leave-One-Subject-Out cross validation show substantial performance increase compared to the state-of-the-arts with UF1 score of 0.98 and UAR score of 1.77. The images attached shows the pre-processing steps done to a sample of the dataset.
Thesis: Lightweight Micro-Expression Recognition
5047981
Data management web tool with Azure components: - Angular pages for editing data - Bulk data upload to Azure - User profile management - All data integration are using .NET core web API - UX/UCD for workflows and page designs - CI/CD for TFS deployment
EMSDB.IO web application
5047965
Ergonomics Assessment Tracking Tool (EATT): - Allows users to submit ergonomics self-assessments - Allows Ergo Contacts to view the submitted self-assessments and input his/her observations and action items - Allows tracking of observations and action items - Dashboard to view critical cases at a glance for Ergo Contact
EATT Web application
10345478
These are data analysis and visualizations I developed for my research that looks at one facet of predicting the spread of Covid-19 cases by looking at the relationship between government policies being put in place and the total number of cases by utilizing Machine Learning techniques.
Data Analysis & Visualization for Covid19 Government Measure
These are data analysis and visualizations I developed for my research that looks at one facet of predicting the spread of Covid-19 cases by looking at the relationship between government policies being put in place and the total number of cases by utilizing Machine Learning techniques.
Data Analysis & Visualization for Covid19 Government Measure
These are data analysis and visualizations I developed for my research that looks at one facet of predicting the spread of Covid-19 cases by looking at the relationship between government policies being put in place and the total number of cases by utilizing Machine Learning techniques.
Data Analysis & Visualization for Covid19 Government Measure
These are data analysis and visualizations I developed for my research that looks at one facet of predicting the spread of Covid-19 cases by looking at the relationship between government policies being put in place and the total number of cases by utilizing Machine Learning techniques.
Data Analysis & Visualization for Covid19 Government Measure
10305823
This is a Flask App for analysing Malay Language Morphology. The main advantage of this analyzer is the ability to morph a word into a noun or verb by adding prefixes to the word. Other advantages of the analyzer I created is that it considers edge cases like the word ‘ajar’ that has different prefix to use (‘pe-‘) and one syllable word like ‘bom’ that uses ‘menge-‘ and ‘penge-‘ prefix. This analyzer is also able to change a word into a intransitive verb e.g. ‘berlari’ by prefixing ‘ber-‘ to the input word.
NLP: Malay Language Morphology Analyser
This is a Flask App for analysing Malay Language Morphology. The main advantage of this analyzer is the ability to morph a word into a noun or verb by adding prefixes to the word. Other advantages of the analyzer I created is that it considers edge cases like the word ‘ajar’ that has different prefix to use (‘pe-‘) and one syllable word like ‘bom’ that uses ‘menge-‘ and ‘penge-‘ prefix. This analyzer is also able to change a word into a intransitive verb e.g. ‘berlari’ by prefixing ‘ber-‘ to the input word.
NLP: Malay Language Morphology Analyser
10305803
Ontology Project: Companies' Named Entity Extraction
10305798
The thesis seeks to demonstrate how transfer learning from macro-expression database in a lightweight network can achieve high Micro-Expression Recognition (MER). The proposed MER model is developed from truncated EfficientNet-B0 model that is pre-trained on a macro-expression database before fine-tuned on the CDE dataset. A simple algorithm-tuning that manipulates the loss function is suggested to deal with the imbalanced dataset. The resulting model is considerably small with just 867K parameters. Experimental results using Leave-One-Subject-Out cross validation show substantial performance increase compared to the state-of-the-arts with UF1 score of 0.98 and UAR score of 1.77. The images attached shows the pre-processing steps done to a sample of the dataset.
Thesis: Lightweight Micro-Expression Recognition
The thesis seeks to demonstrate how transfer learning from macro-expression database in a lightweight network can achieve high Micro-Expression Recognition (MER). The proposed MER model is developed from truncated EfficientNet-B0 model that is pre-trained on a macro-expression database before fine-tuned on the CDE dataset. A simple algorithm-tuning that manipulates the loss function is suggested to deal with the imbalanced dataset. The resulting model is considerably small with just 867K parameters. Experimental results using Leave-One-Subject-Out cross validation show substantial performance increase compared to the state-of-the-arts with UF1 score of 0.98 and UAR score of 1.77. The images attached shows the pre-processing steps done to a sample of the dataset.
Thesis: Lightweight Micro-Expression Recognition
The thesis seeks to demonstrate how transfer learning from macro-expression database in a lightweight network can achieve high Micro-Expression Recognition (MER). The proposed MER model is developed from truncated EfficientNet-B0 model that is pre-trained on a macro-expression database before fine-tuned on the CDE dataset. A simple algorithm-tuning that manipulates the loss function is suggested to deal with the imbalanced dataset. The resulting model is considerably small with just 867K parameters. Experimental results using Leave-One-Subject-Out cross validation show substantial performance increase compared to the state-of-the-arts with UF1 score of 0.98 and UAR score of 1.77. The images attached shows the pre-processing steps done to a sample of the dataset.
Thesis: Lightweight Micro-Expression Recognition
5047981
Data management web tool with Azure components: - Angular pages for editing data - Bulk data upload to Azure - User profile management - All data integration are using .NET core web API - UX/UCD for workflows and page designs - CI/CD for TFS deployment
EMSDB.IO web application
5047965
Ergonomics Assessment Tracking Tool (EATT): - Allows users to submit ergonomics self-assessments - Allows Ergo Contacts to view the submitted self-assessments and input his/her observations and action items - Allows tracking of observations and action items - Dashboard to view critical cases at a glance for Ergo Contact
EATT Web application

Commentaires

Modifications enregistrées
Montre1 - 1 sur 1 commentaires
 Filtrer les commentaires par :
5,0
$225,00 USD
Test and edit Python codes
She delivered great work with minimum instruction! Compled all tasks in 1/2 time allowed. Well communication, will hire again!
Python
Machine Learning (ML)
Big Data
NLP
H
Drapeau de Hoang N.
@hngu178
•
il y a 2 ans

Expérience

Advisor

Techylimpics
juin 2022 - Jusqu'à présent
• Advises and oversees on training materials and competition details such as the activities, judging criteria and source materials for the Artificial Intelligence competition track

Graduate student

UKM
févr. 2020 - mars 2022 (2 ans, 1 mois)
• Masters study in Artificial Intelligence with thesis focusing on automatic and lightweight micro-expression recognition model utilizing transfer learning from macro-expression database. The proposed model is based on EfficientNet-B0 model that is truncated with only 867k parameters. LOSO cross validation has surpassed the state-of-the-arts with UF1 score of 0.98 and UAR score of 1.77. • NLP project : N-Gram tokenizer from scratch

Technical Lead

AirAsia
mars 2021 - déc. 2021 (9 mois, 1 jour)
- Lead, managed and groom requirements for the Health Travel Pass project - Worked on enhancing counter check-in application (Angular/ GCP / .NET API) to include vaccination checks etc - Built web application to capture, annotate, store and retrieve travel documents (ReactJS / NodeJS)

Éducation

MSc Artificial Intelligence

Universiti Kebangsaan Malaysia, Malaysia 2020 - 2022
(2 ans)

BE IT (Software Engineering)

University of Melbourne, Australia 2005 - 2010
(5 ans)

Contactez Nur Aishah A. concernant votre travail

Connectez-vous pour discuter de tous les détails via la messagerie.

Vérifications

Freelance préféré
Identité vérifiée
Paiement vérifié
Téléphone vérifié
E-mail vérifié
Connecté à Facebook

Certifications

c-sharp-prog-1.png C# Programming 1 90%
us_eng_1.png US English 1 85%

Meilleures compétences

Python 1
Machine Learning (ML) 1
.NET
Cloud Computing
C# Programming

Parcourir les freelances similaires

Python Developers in Malaysia
Python Developers
Machine Learning Experts
.NET Developers

Parcourir les présentations similaires

Python
Machine Learning (ML)
.NET
Cloud Computing
Utilisateur précédent Utilisateur suivant
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 FreelanceCatégoriesProjetsConcours FreelancesEntrepriseAdhésionProgramme de Freelance PréféréGestion de projetEmplois locauxPhoto AnywherePrésentationAPI pour les développeursDevenez VérifiéApplication Desktop À proposÀ propos de nousComment cela fonctionneSécuritéInvestisseurPlan du siteHistoiresNouveautésÉquipeRécompensesCommuniqué de presseOffres d'emploi TermesPolitique de ConfidentialitéTermes et ConditionsPolitique de CopyrightCode de conduiteFrais et chargesPartenairesEscrow.comLoadshiftWarrior Forum ApplicationsLogo 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.