Deep learning is a type of machine learning and artificial intelligence (AI) that imitates the way humans gain certain types of knowledge. Deep learning is an important element of data science, which includes statistics and predictive modeling.
In this blog, we will see various of Deep Learning techniques and see how these models work at a high level.
How Neural Network Works?
Before explaining this, let me tell you about some basic terminologies that I am going to use.
- Neurons: Just like human brains, neural networks also contain neurons, where all the data flow and computations happen.
- Hidden Layers: It is a layers between input features and output. It takes the weight from previous layers/input features and produces some output accordingly.
- Features: The features are the elements of your input vectors. The number of features is equal to the number of nodes in the input layer of the network. If you were using a neural network to classify people as either men or women, the features would be things like height, weight, hair length etc.
- Activation Function: It is the most important feature in Deep Learning Neural Networks, without which, the model is equivalent to the linear regression model. It decides the output based on previous input. So, it chooses whether to activate a particular node or not. Let me explain this by taking an example. You touched a hot cup by mistake from your left hand, the nodes, which are in the path from brain to hand will get activated so that you can pull your hand away from the cup. Right-hand neurons won’t get activated. Similar is the case in DNN, activation function decides whether to activate a particular neuron or not.
- Optimizers: There are algorithms or methods used to minimize an error function(loss function) or to maximize the efficiency of production. Optimizers are mathematical functions which are dependent on model's learnable parameters i.e Weights & Biases.
- Loss Function: It calculates the distance between actual output and the predicted one, which helps to improve our model by decreasing the loss.
WORKING:
- The Information(for example, height, weight, etc) is fed into the input layer which transfers it to the hidden layer.
- The interconnections between the two layers assign weights to each input randomly.
- A bias added to every input after weights are multiplied with them individually.
- The weighted sum is transferred to the activation function
- The activation function determines which nodes it should fire for feature extraction.
- The model applies an application function to the output layer to deliver the output
- Weights are adjusted, and the output is back-propagated to minimize error.
- The model uses a cost function to reduce the error rate. You will have to change the weights with different training models.
- The model compares the output with the original result
- It is the process to improve accuracy
- The model adjusts the weights in every iteration to enhance the accuracy of the output.
Types of Deep Learning Networks
Following are the different Deep learning networks:
- Feed Forward Neural Networks: In this type of neural network, flow occurs in the forward direction from input layer to output layer, and hence no back propagation happens. They have only single layer or 1 hidden layer.
- Convolution Neural Networks: It is mostly used in image recognition. This type of network are having more than 3 layers and its used to classify the data which is not linear. These kinds of networks are fully connected with every node.
- Recurrent Neural Networks: RNN is a type of neural network where the output of a particular neuron is fed back as an input to the same node. his method helps the network to predict the output. This network is useful in maintaining a small state of memory which is very useful for developing the chatbot. It is used in chatbot development and text-to-speech technologies.
This was the whole high level stuff you should know. Hope you like this blog. Like and Share!
