Deep Learning with TensorFlow

Deep Learning is the most recent buzzword – which covers most of the Research in todays world. Starting from Image Processing, Vision, Social Network Analysis, Text Analysis, Deep Learning – had taken the show.

This blog mainly covers:
1. Introduction of Artificial Intelligence
2. Relation of Machine Learning with Deep Learning
3. Artificial Neural Networks
4. ML Vs DL
5. DL Neural Network Architecture
6. What is TensorFlow?

Artificial Intelligence

As the human – have deductive reasoning, inference and decision-making so does the computer which is still a long time away, but One thing that have been remarkable gains in the application of AI techniques and associated algorithms.<img

A Deeper Dive into Deep Learning
The Input data is transformed throughout the layers of a deep learning neural network by artificial neurons or processing units. The chain of transformations that occur from input to output is known as the Credit Assignment Path (CAP).

The CAP value is the measurement or concept of ‘depth’ in a deep learning model architecture. According to Wikipedia, most researchers in the field agree that deep learning has multiple nonlinear layers with a CAP greater than two, and some consider a CAP greater than ten to be very deep learning.

Some of the different deep-learning model architectures and learning algorithms include:

  • Feed-forward neural networks
  • Recurrent neural network
  • Multi-layer Perceptrons (MLP)
  • Convolutional neural networks
  • Recursive neural networks
  • Deep belief networks
  • Convolutional deep belief networks
  • Self-Organizing Maps
  • Deep Boltzmann machines
  • Stacked de-noising auto-encoders
  • It’s worth pointing out that due to the relative increase in complexity, deep learning and neural network algorithms can be prone to Over-fitting. In addition, increased model and algorithmic complexity can result in very significant computational resource and time requirements.

    It’s also important to consider that solutions may represent local minima as opposed to a global optimal solution. This is due to the complex nature of these models when combined with optimization techniques such as gradient descent.

    Features of Java Programming Language

    1. Java is PURE-Object Oriented Language:

    Java is object oriented programming language and it religiously follows all OOPS concepts. We cannot write a Java program without creating a class. In order to make the Java program able to run it must have at least one class.

    2. More Security

    Java is a good choice for enterprise web applications due to its security features. In Java there is no concept of user accessible pointers like C or C++. It has been excluded to make untoward modification of addresses and it gives Java security.

    3. Java is Dynamic:

    Java can load a host of libraries at run time which makes Java dynamic. It supports method overriding where the functions corresponding to a specific call are associated at run time.

    Why Java Is Secure Language

    Reason: 1. As java program runs inside its own virtual machine sandbox
    2. Data hiding in Java(OOPs) makes it one of the secure language. Maybe some points are also there but cannot recall it right now.

    3.No use of pointers preventing unauthorized access to memory block.

    4. No access to the memory management</strong>

    Access Control Functionality
    Exception Handling
    Use of final keyword
    Package provides the classes and interfaces for the security framework

    Java security is enabled in each stage:
    1. final keyword secure class loading
    2. jit security manager –  byte code verification automatic memory management
    3. jsse java cryptographic extension digital signature
    4. jaas :java authentication and authorization service

    Arrays or Linked Lists?

    An array implementation of a collection requires O(n) time to search it (assuming it’s not ordered). A linked list also requires O(n) time to search. Yet one of these will be quite a bit faster on a high-performance modern processor. Which one? Why?