Naive Bayes Algorithm for Decision Making

Algorithm used in statistics

1. Statistics and Probability

2. Why you should learn Statistics and Probability Probability and Statistics form the basis of Data Science. The probability theory is very much helpful for making the prediction. Estimates and predictions form an important part of Data science. With the help of statistical methods, we make estimates for the further analysis.

3. Applications of Probability in Real Life Probability is a mathematical term for the likelihood that something will occur. It is the ability to understand and estimate the likelihood of any different combination of outcomes. Let’s discuss some real-life examples of Probability: Weather Forecasting: Before planning for an outing or a picnic, we always check the weather forecast. Suppose it says that there is a 60% chance that rain may occur. Do you ever wonder from where this 60% come from? Meteorologists use a specific tool and technique to predict the weather forecast. They look at all the other historical database of the days, which have similar characteristics of temperature, humidity, and pressure, etc. And determine that on 60 out of 100 similar days in the past, it had rained.

4.   Applications of Probability in Real Life Batting Average in Cricket: Batting average in Cricket represents how many runs a batsman would score before getting out. For example, if a batsman had scored 40 runs out of 100 from boundaries in the previous match. Then, there is a chance that he would score 40% of his runs in the next match from boundaries.  Politics: Many politics analysts use the tactics of probability to predict the outcome of the election’s results. For example, they may predict a certain political party to come into power; based on the results of exit polls. Flipping a coin or Dice: Flipping a coin is one of the most important events before the start of the match. There is no surety, either head will come or not. Both head and tail have 1 out of 2, i.e., 50% chances to occur. Hence, the probability of getting the desired outcome is 0.5. Similarly, while playing with dice, there are 1 out of 6 chances, that the required number will come.

5. Applications of Probability in Real Life Insurance: Probability helps in analyzing the best plan of insurance which suits you and your family the most. For example, you are an active smoker, and chances of getting lungs disease are higher in you. So, instead of choosing an insurance scheme for your vehicle or house, you may go for your health insurance first, because the chance of your getting sick are higher. For instance, nowadays people are getting their mobile phones insured because they know that the chances of their mobile phones getting damaged or lost are high. Playing Cards: There is a probability of getting a desired card when we randomly pick one out of 52. For example, the probability of picking up an ace in a 52 deck of cards is 4/52; since there are 4 aces in the deck. The odds of picking up any other card is therefore 52/52 – 4/52 = 48/52. Lottery Tickets Are we likely to die in an accident?

6. Probability is the Bedrock of Machine Learning Classification models must predict a probability of class membership. Algorithms are designed using probability (e.g. Naive Bayes). Learning algorithms will make decisions using probability (e.g. information gain). Sub-fields of study are built on probability (e.g. Bayesian networks). Algorithms are trained under probability frameworks (e.g. maximum likelihood). Models are fit using probabilistic loss functions (e.g. log loss and cross entropy). Model hyperparameters are configured with probability (e.g. Bayesian optimization). Probabilistic measures are used to evaluate model skill (e.g. brier score, ROC).

7. Basic Probability

8. Joint Probability

9. Joint Probability

10. Marginal Probability Distribution

11. Probability Example

12. Probability Example

13. Probability Example

14. Probability Example (Not Joint Probability)

15. Conditional Probability

16. Conditional Probability

17. Conditional Probability

18. Conditional Probability

19. Naïve Bayes Classifier Day Outlook Temperature Routine Wear Coat? D1 Sunny Cold Indoor No D2 Sunny Warm Outdoor No D3 Cloudy Warm Indoor No D4 Sunny Warm Indoor No D5 Cloudy Cold Indoor Yes D6 Cloudy Cold Outdoor Yes D7 Sunny Cold Outdoor Yes

20. Naïve Bayes Classifier (Test Set) Day Outlook Temperature Routine Wear Coat? D1 Cloudy Warm Outdoor ??

21. Naïve Bayes Classifier (Test Set) Day Outlook Temperature Routine Wear Coat? D1 Cloudy Warm Outdoor ?? P(C=Yes Wear Coat|X=Cloudy, warm, outdoor) and P(C=No Wear Coat|X=Cloudy, warm, outdoor)

22. Naïve Bayes Classifier (Test Set-Case I)  

23. Naïve Bayes Classifier (Test Set Case II)   Day Outlook Temperature Routine Wear Coat? D1 Cloudy Warm Outdoor No

24. Thank You