# Artificial intelligence machine learning and data science

## Types of Machine Learning Algorithms

Machine Learning algorithms are classified as – 1. Supervised Machine Learning AlgorithmsMachine learning algorithms that make predictions on given set of samples.Supervised machine learning algorithm searches for patterns within the valuelabels assigned to data points. 2. Unsupervised Machine Learning AlgorithmsThere are no labels associated with data points. These machine learningalgorithms organize the data into a group of clusters to describe itsstructure and make complex data look simple and organized for analysis. 3. Reinforcement Machine Learning AlgorithmsThese algorithms choose an action, based on each data point and later learnhow good the decision was. Over time, the algorithm changes its strategy tolearn better and achieve the best reward. Common Machine Learning Algorithms Infographic

## Difference Between Data Science and Machine Learning

Data science is all about understanding business problems using data. Itstarts with collecting data, then understanding the data, cleansing andtransforming it, and then we make decisions about the future course of action.Machine learning is like an afterthought after the full process of datascience.> Also, Read – 100+ Machine Learning Projects Solved and Explained.But one thing to understand here is that ML is one of the most essentialthoughts in the entire process of data science.So when we are working with data to make future decisions on a businessstrategy, it is data science. So far you must have some idea that ML is partof data science, but why do we have a separate term for it? If ML is a conceptof data science, why is it just as important as data science?

## Data Science Vs. Machine Learning: Examples

The main difference between data science and machine learning is that in datascience we study the problems of a business and make decisions based on ourobservations, where ML is used to create and use models that can learn fromthe data. Simply put, machine learning can also be called predictivemodelling.In Data Science we analyze a business using data, but in ML our main goal isto create models to predict future outcomes on new data. Some of the examplesof using machine learning models are as follows: 1. Predict whether an email is spam or not 2. Predict fraud transactions 3. Predicting which ad to show to which user 4. Predicting the next world cup winner.Now let’s take a look at examples of the same business model to see where weuse data science and where ML. The popular dating app “Tinder” collects datafrom its members to find the most suitable match for them. At this point,Tinder is using data science. But if we start analyzing the data to predicthow okay someone is to sleep with you on the first date, then at this pointTinder is using machine learning.

## Artificial intelligence, machine learning, and data science

Artificial intelligence, machine learning, and data science are often usedinterchangeably. Actually, they are different but overlapping domains. As Ialready noted, artificial intelligence has a broader scope than machinelearning. Machine learning is just one facet of artificial intelligence.Similarly, some argue that data science is a facet of artificial intelligence.Others say the opposite, that data science includes AI.In the field, data scientists and AI experts offer different kinds ofexpertise with some overlap. Data science uses many machine learningalgorithms, but not all of them. The Venn diagram in Figure 1 shows the spaceswhere artificial intelligence, machine learning, and data science overlap.Figure 1: The overlaps between artificial intelligence, machine learning, anddata science.Note: See Data Science vs. Machine Learning and Artificial Intelligence formore about each of these technology domains and the spaces where they meet.

## Choosing a machine learning algorithm

We now have a typical classification problem: Given the incoming data, thealgorithm must find a class for those data. In other words, it has to labeleach data item approved or denied. Because we have the manager’s collectedresponses, we can use a supervised learning method. We only need to choose thecorrect algorithm. The major machine learning algorithms are: * Linear Regression * Logistic Regression * K-Nearest Neighbors * Support Vector Machines * Decision Trees and Random Forests * Neural NetworksNote: For more about each of these algorithms, see 9 Key Machine Learning Algorithms Explained in Plain English.Except for linear regression, we could apply any of these algorithms to ourclassification problem. For this use case, we will use a Logistic Regressionmodel. Fortunately, we don’t need to understand the algorithm’s implementationdetails. We can rely on existing tools for implementation.

## Data Science vs. Machine Learning

* * *Data science is a process of extracting information from unstructured/rawdata. To accomplish this task, it uses several algorithms, ML techniques, andscientific approaches. Data science integrates Statistics, Machine Learning,and Data Analytics. Below we are narrating 15 distinctions between DataScience vs. Machine Learning. So, let’s start.

## Machine learning in Data Science

To become a data scientist, one should also be aware of machine learning andits algorithms, as in data science, there are various machine learningalgorithms which are broadly being used. Following are the name of somemachine learning algorithms used in data science: * Regression * Decision tree * Clustering * Principal component analysis * Support vector machines * Naive Bayes * Artificial neural network * AprioriWe will provide you some brief introduction for few of the importantalgorithms here,1. Linear Regression Algorithm: Linear regression is the most popular machinelearning algorithm based on supervised learning. This algorithm work onregression, which is a method of modeling target values based on independentvariables. It represents the form of the linear equation, which has arelationship between the set of inputs and predictive output. This algorithmis mostly used in forecasting and predictions. Since it shows the linearrelationship between input and output variable, hence it is called linearregression.The below equation can describe the relationship between x and y variables:Where, y= Dependent variable X= independent variable M= slope C= intercept.2. Decision Tree: Decision Tree algorithm is another machine learningalgorithm, which belongs to the supervised learning algorithm. This is one ofthe most popular machine learning algorithms. It can be used for bothclassification and regression problems.In the decision tree algorithm, we can solve the problem, by using treerepresentation in which, each node represents a feature, each branchrepresents a decision, and each leaf represents the outcome.Following is the example for a Job offer problem:In the decision tree, we start from the root of the tree and compare thevalues of the root attribute with record attribute. On the basis of thiscomparison, we follow the branch as per the value and then move to the nextnode. We continue comparing these values until we reach the leaf node withpredicated class value.3. K-Means Clustering: K-means clustering is one of the most popularalgorithms of machine learning, which belongs to the unsupervised learningalgorithm. It solves the clustering problem.If we are given a data set of items, with certain features and values, and weneed to categorize those set of items into groups, so such type of problemscan be solved using k-means clustering algorithm.K-means clustering algorithm aims at minimizing an objective function, whichknown as squared error function, and it is given as:Where, J(V) => Objective function ‘||xi – vj||’ => Euclidean distance between xi and vj. ci’ => Number of data points in ith cluster. C => Number of clusters.* * *

## How to solve a problem in Data Science using Machine learning algorithms?

Now, let’s understand what are the most common types of problems occurred indata science and what is the approach to solving the problems. So in datascience, problems are solved using algorithms, and below is the diagramrepresentation for applicable algorithms for possible questions:Is this A or B? :We can refer to this type of problem which has only two fixed solutions suchas Yes or No, 1 or 0, may or may not. And this type of problems can be solvedusing classification algorithms.Is this different? :We can refer to this type of question which belongs to various patterns, andwe need to find odd from them. Such type of problems can be solved usingAnomaly Detection Algorithms.How much or how many?The other type of problem occurs which ask for numerical values or figuressuch as what is the time today, what will be the temperature today, can besolved using regression algorithms.How is this organized?Now if you have a problem which needs to deal with the organization of data,then it can be solved using clustering algorithms.Clustering algorithm organizes and groups the data based on features, colors,or other common characteristics.* * *

## Data science / Machine learning

* Dataquest — Teaches you Python and data science interactively. You analyze a series of interesting datasets ranging from CIA documents to NBA player stats. You eventually build complex algorithms, including neural networks and decision trees. * Python for Data Analysis — written by the author of a major Python data analysis library, it’s a good introduction to analyzing data in Python. * Scikit-learn documentation — Scikit-learn is the main Python machine learning library. It has some great documentation and tutorials. * CS109 — this is a Harvard class that teaches Python for data science. They have some of their projects and other materials online.

## Data Science / Machine Learning Project Ideas

* A map that visualizes election polling by state. * An algorithm that predicts the weather where you live. * A tool that predicts the stock market. * An algorithm that automatically summarizes news articles.You could make a more interactive version of this map. From RealClearPolitics.

## Using the Right Machine Learning Algorithm

Machine learning algorithms can only understand and learn the training datasets and its use. If you choose the inapplicable algorithm your AI projectwill be failed or will give inaccurate results, which has no room in thehealthcare sector.So, train your model with efficient dental image analysis algorithms as perthe availability of types of data and model validation system. And to improvethe machine learning algorithm performance, a huge amount of data is required.Also Read: How Much Training Data is Required for Machine Learning Algorithms?

## Data Mining, Machine Learning Vs Data science

[image source]Data Science is a vast area under which Machine Learning comes. Manytechnologies such as SPARK, HADOOP, etc also come under data science. Datascience is an extension of statistics which has the capability to processmassively large data using technologies.It deals with all real-world complex problem solving such as requirementanalysis, understanding, extracting useful data, etc.Data Science deals with human-generated raw data, it can analyze the images,and audios from data just like how humans do. Data science requires a highskill set with domain expertise, strong knowledge of databases, etc. Itdemands high computational resources, high RAM, etc.Data Science models have clearly defined milestones to achieve when comparedto Machine Learning which tries to achieve the target only with the availabledata.Data Science Model comprises of: * ETL- Extract Load and Transform data. * Data Distribution and processing. * Automated models application for outcomes. * Data Visualization * Reporting with slice and dice feature for better understanding. * Data Backup, recovery and security. * Migration to production. * Running business models with the algorithms.

## Machine learning algorithms:

* Decision tree learning * Artificial neural networks * Deep Learning * Association rule also in learning * Support vector machines * Inductive logic programming * Reinforcement also learning * Genetic algorithm * Sparse dictionary also in learning * Bayesian networks