What Topic Should I Cover Next?
Choose one of the four options or comment something else!
Hello everyone, it has been a while since my last post; but I have some free time to make some more in depth articles over deep learning and artificial intelligence.
I have a few topics I could cover, but I want to you, the viewers, to decide on which exactly I should start with next! Here are the four possible options:
- Complete Guide to Deep Learning Activation Functions
- Complete Guide to Deep Learning Optimizers
- Efficient Neural Architecture Search using a Genetic Algorithm
- Data Augmentation Techniques for Continuous Data
Complete Guide to Deep Learning Activation Functions (1) will cover all recent advancements in deep learning activation functions, their improvements and intuitions. Beginning from simple sigmoid and tanh, to the rectified linear unit family of ReLU, Leaky ReLU, ELU, SeLU, GeLU; up to newer functions like Swish and Mish!
Complete Guide to Deep Learning Optimizers (2) will cover recent advancements in deep learning optimizers, their improvements and intuitions. Beginning from simple SGD, we will work our way up to Adam through RMSProp and Adagrad; eventually covering newer optimizers such as AdamW, Demon Adam, and Power Sign! In addition, we will be implementing these optimizers from scratch!
Efficient Neural Architecture Search using a Genetic Algorithm (3) will cover my thesis research in the realm of neural architecture search (NAS) using a genetic algorithm. NAS refers to the process of artificially finding the architecture of a deep learning model; in this case a convolutional neural network!
Data Augmentation Techniques for Continuous Data (4) will cover a few simple but yet powerful techniques for augmenting continuous datasets. When the dataset size is limited, being able to augment data samples in order to prevent model overfitting is vital for performance.
Now that you know which topics are available and what they concern, please comment the number corresponding to the topic, and in two weeks or so I’ll select the topic with the most votes/comments for the next article! If you have a suggestion for a future article that lines up with what I have done before, please comment that as well and I will consider it for the future!
If you are looking for more information on artificial intelligence and machine learning, check out these two other posts!
