Hello Shima,

You can use a neural network for these two cases (when an image is input, you have to convert it to a sequence of pixel intensities). Neural networks are universal. If you want better quality of predictions you can use random forest for sequence and convolution networks for image.

You can read more in Christopher M. Bishop's book "Pattern Recognition and Machine Learning"

The site https://scikit-learn.org/ has many implementations of ML techniques in Python.

Regards,

Georgy Kossov

Dear Shima Shafiee ,

n particular, convolutional neural networks (CNN) have achieved great success for image-based tasks, and recurrent neural networks (RNN) are renowned for sequence-based problems. Specifically, deep learning methods based on the CNN and RNN architectures have been adopted for motion recognition using RGB-D data.

https://www.sciencedirect.com/science/article/abs/pii/S1077314218300663#:~:text=In%20particular%2C%20convolutional%20neural%20networks,recognition%20using%20RGB-D%20data.

Regards,

Shafagat

For sequence based data, RNN, LSTM works great. Transformers are great for sequenced based data as well but needs a lot for data to train unless you use a pretrained model. You have to find out which works best for the type of data you are using and the kind of task you want to do.

DEAR Mosarrat Rumman, thanks for your guidance

DEAR Aparna Sathya Murthy

The input data of the deep model is one-dimensional and a string of amino acids. Many thanks for sharing the relevant references.

DEAR Georgy Kossov

Thank you for your support and guidance

DEAR Shafagat Mahmudova

Thank you for your support and guidance

Hello Shima Shafiee :

In particular, convolutional neural networks (CNN) have achieved great success for image-based tasks, and recurrent neural networks (RNN) and LSTM are renowned for sequence-based problems. You have to find out which works best for the type of data you are using and the kind of task you want to do.

Regards;

Ehtisham

DEAR Ehtisham Lodhi,thanks for sharing your idea