Python is strong candidate in this field but if you want speedy work without programming or if you dont have a good programming Language knowledge than go for MATLAB or OCTAVE softwares.
All language is better than each other. Then it depends on you which language you known and comfortable. If you are convenient in Python language you do.
All languages are pretty good for deep learning as long as the underlying concepts of techniques are understood. It however depends on which language is one comfortable with. A lot of ready made codes seem available online for python users and a lot of libraries are there. Using those ready made codes may prevent one from having the real knowledge if care is not taken.
Currently, pytorch (supported by facebook) and tensorflow (supported by google) are the most popular deep learning frameworks. Both of them are based on python. Pytorch are more popular among reserachers and tensorflow are more popular among industrial people.
If you're aiming at Deep Learning, the answer is definitely Python. Because the very powerful Tensorflow (Keras) and Pytorch. And most of the open source deep learning code online is written with Python, which means you have an easy access to interesting and helpful programs in this field writen with Python. And if you take a look at the online open courses about deep learning offered by a lot of famouse universities, you will find most of them are using Python.
The Python ecosystem is growing and may become the dominant platform for machine learning. The primary rationale for adopting Python for ML & DL is because it is a general purpose interpreted and complete language and platform (unlike Matlab or R ). This means that you can use the same code for research and development to figure out what model to run as you can in production. thus, The cost and maintenance efficiencies and benefits of this fact cannot be understated. further, there are also a lot of modules and libraries to choose from what you want to be suit your task, providing multiple ways to do each task.
If you are a beginner then have a look at MATLAB. It has a lot of examples but you must learn python for your research work because it has vast libraries that you may need to complete your work.
Start with Python as the base programming language and Keras, Tensorflow as deep learning frameworks. You can easily move to Pythorch or MxNet or any other DL frameworks if you are not comfortable with Keras.
For the beginning and understand, Matlab is very convenient but furthermore, for projects and industrial applications Python, Tensorflow are more suitable.
Dear Ridwan, in general python is a good choice for simplicity and versatility, in addition to having a large number of libraries to work with deep learning
I think the most challenging task in machine learning and deep learning is the task of preprocessing your Dataset.
so before even thinking about the way of applying your model of interest, you should be worried about the data, but don't be worry! because python a lot of package in order to assist you in that manner. python has a lot of great and useful packages that can give you a hand of help.
so in my openion it is more than an applying a model, and you should look at this task in that way too!
Python and/or R. I prefer R for preprocessing/analysis but I increasingly see the appeal of performing the modelling in Python and the field seems to be moving in that direction. MATLAB also seems interesting, but it's hard to beat free.
I agree with the other answers that Python is the most convenient approach for general applications and tutorials. However, for a more practical "what´s best" answer you'd always need to consider the concrete use case. If you know what you would like to do you may look for a framework providing possible solutions/libs for that specific application. When you've found a framework capable of dealing with your topic, most likely there is already some code available you could start to work with. So, my advice: Stay flexible in terms of languages and frameworks and make sure to fully understand the demands of your problem at first.