PHD Discussions Logo

Ask, Learn and Accelerate in your PhD Research

Question Icon Post Your Answer

Question Icon

6 years ago in Data Analysis By Trisha

Limitations of artificial neural networks

All Answers (4 Answers In All)

By Pranav Answered 6 years ago

On a technical level, the biggest challenge is the amount of time taken to train the networks, 
which require a great amount of power for performing more complex tasks. 
The biggest issue, is that the artificial neural networks are black boxes. 
Here the user can feed in data and can receive an answer. The user can fine-tune the answers, but they don’t have any access to decision making process.

By Veena Answered 6 years ago

Yes there are a few limitations which are being worked upon. https://arxiv.org/pdf/1404.5598.

By Adi Answered 6 years ago

neural networks are designed for spotting data patterns. Some of the tasks include classification, clustering, and prediction.

By Sourabh Answered 6 years ago

Artificial neural networks can perform a number of tasks. From making cars drive autonomously to machine translation. 
Generating realistic CGI faces, fraud detection, reading our minds, etc. the neural nets lead to many of the biggest advances in artificial intelligence.

Your Answer