Machine Learning in Depth

What is Big Data ? Nowadays Big data is of great importance. It is usually based on an idea that we need not manage data in separate chunks or bundles. Instead, we keep all sorts of data in a warehouse like server. Only useful information is extracted from the warehouse called Data Mining. In order to be expert in Machine Learning, You should be familiar with Data Mining also. Mining is simply known as extracting useful information from unnecessary materials. We should be able to separate unuseful information with useful information. Parametric and Nonparametric Machine Learning algorithm. Examples of parametric machine learning algorithms are Linear Regression, Logistic Regression Non-parametric Examples of Machine Learning algorithms are Suppport Vector Machines, Neural Networks and Decision Trees.

First, follow the link below.You will come across a lot of errors. Just bare with me. https://www.tensorflow.org/install/install_windows If you are using anaconda with python 3.6 you should type these to make tensorflow work:: conda create --name tensorflow python= 3.5 activate tensorflow conda install jupyter conda install scipy pip install tensorflow \# or \# pip install tensorflow-gpu because currently only python 3.5 supports tensorflow or has no problem. Tensorflow is one of the emerging software in Machine Learning. Using tensorflow we can define classifiers and train those to perform various applications. FAQ. Where to type python commands? Python commands can be typed in command prompt. First you need to activate python in command prompt by typing python command. FAQ. How to activate tensorflow in command promt? We use command activate tensorflow. FAQ. What is Anaconda? Anaconda is just a complex platform which contains all tensorflow,...

Bayesian reasoning(Naive Bayes Algorithm) This algorithm is one of the important algorithm in Machine Learning. Many people make mistake in this. Bayesian Reasoning can be very important for any field which includes Statistics. Lets try to grab the concept. Bayesian reasoning is logical reasoning.  Even doctors make this mistake. That can be very dangerous. Question: Imagine a women in her forties and she wants to know if she has breast cancer. If she does, statistics show that there is 75% chance that a mammogram will detect it but if she doesn’t there is a 10% chance that the mammogram will give her a false positive, that is, it will say she has cancer and she doesn’t. It is a fact that 1.4% of all women in their forties  have breast cancer. Now imagine a hypothetical Women gets a mammogram and it comes back positive. What are the odds that the test is right? Answer: First of all change the percent to per thousand. 1.4 % means out of 1000 people 14 women has...

Supervised learning stands for training a computer or a machine or we can say feeding the data to an algorithm so that the program can have a close to accurate guess. If you want to distinguish between apple and orange, only the color and texture is not sufficient. We need to first train the program with training data set and this helps to analyze all the texture, color and pattern of the all the different pictures of the training data set and can help to recognize the fruit to a higher degree though not accurate. Example: Facebook  uses your photos from your profile or the tagged photos and trains the program thus training data set and can recognize you from the photo given. Another example would be if you can extract the information and import it to the program about the temp and genre of the music you like. Then it can definitely recommend a new song. Stock markets can also be analyzed from the same method so this is a very useful method. Example of non-supervised l...

Syllabus: These are the topics you need to master in order to learn Machine Learning. These courses are available readily in various internet sites and also best Harvard lectures by Andrew Ng is also available online. The syllables of Machine Learning are: 1. Introduction 2. Linear Regression with One Variable 3. Linear algebra 4. Linear Regression with Multiple Variables 5. Octave/Matlab 6. Logistic Regression 7. Regularization 8. Neural Networks: Representation 9. Neural Networks: Learning 10. Support vector machine 11. Unsupervised Learning 12. Dimensionality Reduction 13. Anomaly Detection 14. Recommender Systems 15. Large scale Machine Learning 16. Application: Photo OCR So these are the vast courses that you need to learn in order to master in Machine Learning. Don’t get frustrated at first because good things take time. This might look very difficult but if you have that will power that I can do it and I will do it at any costs, bel...