Abstracts Engineering

Add abstract

Want to add your dissertation abstract to this database? It only takes a minute!

Search abstract

Search for abstracts by subject, author or institution

Share this abstract

Building and Training a Fully Connected Deep Neural Network From Scratch

by Axel Berglund

Institution: KTH
Department: Electrical Engineering and Computer Science (EECS)
Degree:
Year: 2022
Keywords: Deep Neural Network; Machine Learning; Gradient Decent; MNIST.; Electrical Engineering, Electronic Engineering, Information Engineering; Elektroteknik och elektronik
Posted: 3/25/2025
Record ID: 2267521
Full text PDF: http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-323615


Abstract

Artificial Neural Networks make up the core of mostMachine Learning algorithms. In the past decade Machine learninghave successfully taken on fields such as image recognition,Data analytics and medical technologies. As the area of usebecome less prone to mistakes, it raises the responsibility lookinto the black box of code and understand it to a deeper level. Inthis project, I built a Deep Neural Network from scratch, withouthigh level libraries, and trained it for a supervised classificationtask. The finished algorithm is flexible and can be adapted toany classification problem. The training method is based onBackpropagation and Gradient Descent. At last, the algorithmwas trained on the Modified National Institute of Standardsand Technology (MNIST) database, and performed with a 77%prediction acccuracy. There are a few optimization methods yetto be tested to further increase the performance. Artificiella neurala nätverk utgör kärnan i de flesta maskininlärningsalgoritmer idag. Under det senaste decenniet har maskininlärning framgångsrikt tagit an områden som bildigenkänning, dataanalys och medicinsk teknik. När användningsområdena blir mindre benägna till misstag, ökar ansvaret av att titta under huven och förstå den djupare nivåkoden. I denna studie var syftet att bygga ett djupt neuralt nätverk från grunden, utan högnivåbibliotek, och träna det för en övervakad klassificeringsuppgift. Den färdiga algoritmen är flexibel och kan designas för flera klassificeringsproblem. Nätverkets träningsmetod är baserad på Backpropagation och Gradient Descent. Valideringsdatan kunde till slut köras med 77% korrekt noggrannhet, och det finns finns ytterligare optimeringsmetoder att testa för att höja prestationen. Kandidatexjobb i elektroteknik 2022, KTH, Stockholm

Add abstract

Want to add your dissertation abstract to this database? It only takes a minute!

Search abstract

Search for abstracts by subject, author or institution

Share this abstract

Relevant publications

Book cover thumbnail image
Predicting the Admission Decision of a Participant...
by Yigit Ozsert, Gozde
   
Book cover thumbnail image
Development of New Models Using Machine Learning M...
by Akgol, Derman
   
Book cover thumbnail image
The Adaptation Process of a Resettled Community to... A Study of the Nubian Experience in Egypt
by Fahmi, Wael Salah
   
Book cover thumbnail image
Development of an Artificial Intelligence System f...
by Chand, Praneel
   
Book cover thumbnail image
Theoretical and Experimental Analysis of Dissipati...
by Latour, Massimo
   
Book cover thumbnail image
Optical Fiber Sensors for Residential Environments
by García-Olcina, Raimundo
   
Book cover thumbnail image
Calibration of Deterministic Parameters Reassessment of Offshore Platforms in the Arabian ...
by Zaghloul, Hassan
   
Book cover thumbnail image
How Passion Relates to Performance A Study of Consultant Civil Engineers
by Cadieux, Trevor J.