What is machine learning and how it is related to artificial intelligence?
Machine learning is a subject in which we train machines to act like humans.
How does machine learning work?
In machine learning we train machine learning models with a set of data and on basis of that data it predicts and gave us result.
We can understand the working way of machine learning model from these steps.
1. Data collection and cleaning
In this step, we collect various datas which is relevant in training of our machine learning model, we can use different sources to collect the data such as databases, internet.
After we collect our valuable data then we need to clean it by fixing the errors in the data, categorisation of the data, removing the corrupted data.
2. Training and testing data
After collecting and Cleaning of our data we will divide it into two parts training and testing data. In training data we take most of the data to give as an input to for model to train on and for testing data we take remaining datas to test the model once it gets trained on our algorithm.
Let's example we have 50 thousand images as data, so we will take 35 thousand data for training and remaining 15 thousand as our testing data.
3. Writing the algorithms
In this step we write the algorithm on which our model would be trained. Let's example if we are training a neutral network on the basis of our data then we will write the algorithms and create different layers of neutral networks to train further.
4. Training
Now, we will start the training of our model.
In supervised learning, the model is trained using labeled data, which means the input data is paired with the corresponding correct output. During training, the model adjusts its internal parameters iteratively to minimize the difference between its predictions and the actual labels.
if our data is very large then it takes longer to train it, it takes hours or even days to train a large Machine learning models, the training period is also depend of computerization power.
5. Testing
After training of our machine learning model, the model's performance is evaluated using a separate set of data called the validation set. This helps to ensure that the model is not overfitting to the training data. After that it is also trained of another dataset and that is testing dataset which is kept initially seperate to training data.
6. Deployment
Once our model passed the test and start working perfectly then it is time to deploy the model in real world to predict on unseen data and solve the problem it was trained for.
Why do we need machine learning?
There are many fields where machine learning is used and without the help of machine learning we would never be reached so high in the field of technology.
1. Pattern recognition
Machine learning models is far more better, efficient and fast than us humans in pattern recognition.
It is used in many works where we need pattern recognition.
A great example of this is a.i. bots which are training to play chess, thay are so accurate that even it becomes hard for top players to compete with them.
2. Automating tasks
Automation of the machines is one of the best field where we get to watch the power and need of machine learning. Without it we would never have any auto driving cars, automated machines which works in factories and many more.
3. Improving efficiency
It improves the efficiency of our work by helping us in many fields, in factories the machines are used in manufacturing of products and increase the speed and efficiency of production not only that there are many machines which helps in agriculture to produce more and better amount of food with less resources.
4. Fraud detection
In the age of internet scammers have discovered many new ways to scam and fraud so machile learning plays an important in fraud detection and prevention from attacks.
5. Personalization
Machine learning helps companies to make their products more personalized to the users such as Facebook, YouTube and many companies uses machine learning algorithm to keep tracks of likes and dislikes of users and suggest their content according to the data.
Types of machine learning?
There are many three types of machine learning.
-Supervised machine learning
-Unsupervised machine learning
-Reinforcement machine learning
Supervised machine learning
In supervised machine learning, we use labelled data for training of the model
meaning that each input data point is associated with a corresponding target or output label. The goal of supervised learning is to learn a mapping function that can accurately predict the output labels for new, unseen input data.
Unsupervised machine learning
Unsupervised machine learning is a type of machine learning in which we only provide datasets which are not labelled in any form, the models automatically learns to read the pattern in the dataset and get trained on its own.
The goal of the unsupervised learning is to find the patterns, relationship in the data on its own without any guidance such as labelled dataset.
Reinforcement machine learning
In reinforcement machine learning we don't provide any datasets to the machine but it automatically get trained according to the environment it is in.
It gets trained by taking actions and observing the consequences of the action taken by it. After taking an action the agent receives signal from the environment in the form or a reward, the reward indicates the goodness of the action and the machine's goal is to maximize the earned reward each time.
Everytime it gets new reaction from the environment it explores some more actions to discover better strategies.
Artificial intelligence vs machine learning?
Artificial intelligence is a broad term in computers and technology it's goal is to create robots and machines which act like humans.
The basic goal of artificial intelligence is pattern recognition, prediction, problem solving, learning, understanding, decision making.
The a.i. also includes traditional ways to train a robot to mimic humans, it involves the use of explicit rules and predefined knowledge to perform specific tasks. These rules are created and programmed by human experts. While they can be effective for certain well-defined tasks, they often struggle with handling complex, unstructured data and require manual updates as the task or domain changes.
On the other side, machine learning is a subset of artificial which mainly focuses on training of the model to recognise patterns, making prediction and decision making on data, it relies on input data to learned from and no on any handcrafted rules
What is neural network?
A neural network is a computational model inspired by the structure and function of the human brain, it is a subset in machine learning.
They are designed to learn to make predictions and analyse pattern from a set of data.
The basic building block of neural network is called neuron, it is inspired from neurons which is inside human brain.
Neurons takes data as input perform a simple computation on that input, and produce an output.
These neurons are organised into layers and information flows through the layers from input to output layers.
History of machine learning?
The idea of machine learning had came in the mind of some of the seekers such as Alan Turing(the inventor of turing machine), who proposed the idea of a "universal machine" capable of simulating any algorithmic process. In the 1950s, the field of artificial intelligence (AI) emerged, and researchers began exploring the idea of creating machines that could mimic human intelligence and perform better in the fields where humans can't.
Then in 1956 the Dartmouth conference was organised which officially launched artificial intelligence.
The Dartmouth Summer Research Project on Artificial Intelligence, held from 18 June through 17 August of 1956, is widely considered the event that kicked off AI as a research discipline. Organized by John McCarthy, Marvin Minsky, Claude Shannon, and Nathaniel Rochester, it brought together a few dozen of the leading thinkers in AI, computer science, and information theory to map out future paths for investigation.
Researchers began developing machine learning algorithms, one of the earliest ml algorithm was "Perceptron" by Frank Rosenblatt in 1957, which laid the foundation for neural networks. However, initial enthusiasm for neural networks waned due to perceived limitations.
During mid 1960s, AI research shifted toward "symbolic AI," which relied on handcrafted rules and explicit representations of knowledge. Expert systems, rule-based systems designed to mimic human expertise in specific domains, became prominent.
Progress in AI and machine learning faced significant challenges, and funding for AI research declined. This period, known as the "AI winter," was characterized by reduced interest and progress in the field.
Enterprise
In 1969, Minsky and another AI researcher, Seymour Papert, published a book called Perceptrons, which pointed out the flaws and limitations of neural networks. This publication influenced DARPA to withdraw its previous funding of AI projects.
In 1973, an evaluation of academic research in the field of AI called the "Lighthill Report" was published. It was highly critical of research in the field up to that point, stating that AI research had essentially failed to live up to the grandiose objectives it laid out. This report caused the U.K. to cease funding for AI. This ushered in the first AI winter, which took place between 1974-1980, after a nearly 20-year period of significant interest during what some have called AI's Golden Era. Interest in AI wouldn't be revived until years later with the advent of expert systems, which used if-then, rule-based reasoning. This would eventually end with another AI winter from the late 1980s to mid-90s.
During 2010s, The explosion of digital data, increased computing power, and advancements in algorithms led to significant progress in machine learning. Deep Learning, a subset of neural networks with multiple layers, emerged as a powerful approach to handle complex data, leading to breakthroughs in areas like image recognition, natural language processing, and speech recognition.
And now at present, a.i. is at most developed form and most successful ever in history and It has a very big potential.
Many big companies such as Microsoft, google etc are working to make it more better.
.png "How to create a registration form using php, html, css and mysql database.")
.png "How to create a registration form using php, html, css and mysql database.")
.png "How to create a registration form using php, html, css and mysql database.")
.jpeg)
.jpeg)
.png)
.png)
