1. What Is Ai?
Answer: Artificial intelligence (“AI”) can mean many things to many people. Much confusion arises that the word ‘intelligence’ is ill-defined. The phrase is so broad that people have found it useful to divide AI into two classes: strong AI and weak AI.
2. What are common uses and/or applications for AI?
Answer: here should show that you recognize the far-reaching and practical applications of AI, but your answer is up to you because your personal understanding of the AI field is what the interviewer is trying to ascertain. If possible, mention those uses most relevant to the potential employer. Possibilities include contract analysis, object detection, and classification for avoidance and/or navigation, image recognition, content distribution, predictive maintenance, data processing, automation of manual tasks or data-driven reporting.
3. What Are Good Programming Languages For Ai?
This topic can be somewhat sensitive, so I’ll probably tread on a few toes, please forgive me. There is no authoritative answer to this question, as it really depends on what languages you like programming in. AI programs have been written in just about every language ever created. The most common seem to be Lisp, Prolog, C/C++, recently Java, and even more recently, Python.
LISP: For many years, AI was done as research in universities and laboratories, thus fast prototyping was favored over fast execution. This is one reason why AI has favored high-level languages such as Lisp. This tradition means that current AI Lisp programmers can draw on many resources from the community. Features of the language that are good for AI programming include garbage collection, dynamic typing, functions as data, uniform syntax, interactive environment, and extensibility. Read Paul Graham’s essay, “Beating the Averages” for a discussion of some serious advantages:
PROLOG: This language wins ‘cool idea’ competition. It wasn’t until the 70s that people began to realize that a set of logical statements plus a general theorem prover could make up a program. Prolog combines the high-level and traditional advantages of Lisp with a built-in unifier, which is particularly useful in AI. Prolog seems to be good for problems in which logic is intimately involved, or whose solutions have a succinct logical characterization. Its major drawback (IMHO).
4. What Is Relational Knowledge?
Answer: It is a knowledge representation scheme in which facts are represented as a set of relations. For example knowledge about players can be represented using a relation called “player” having three fields: player name, height, and weight. This form of knowledge representation provides weak inferential capabilities when used as standalone but are useful as an input for sophisticated inferential procedures.
5. What Are Frames And Scripts In “Artificial Intelligence”?
Answer: Frames are a variant of semantic networks which is one of the popular ways of presenting non-procedural knowledge in an expert system. A-frame which is an artificial data structure is used to divide knowledge into substructure by representing “stereotyped situations’. Scripts are similar to frames, except the values that fill the slots must be ordered. Scripts are used in natural language understanding systems to organize a knowledge base in terms of the situation that the system should understand.
6. What is Tensorflow?
Answer: TensorFlow is an open source machine learning framework. It is fast, flexible and a low-level toolkit meant for doing a complex algorithm and offers customizability to build experimental learning architectures. AlphaGo and Google Cloud Vision are built on Tensorflow framework.
7. When will you use classification over regression?
Answer: Classification is used when the output variable is a category such as “red” or “blue”, “spam” or “not spam”. It is used to draw a conclusion from observed values. Differently from, regression which is used when the output variable is a real or continuous value like “age”, “salary”, etc. When we must identify the class, the data belongs to we use classification over regression. Like when you must identify whether a name is male or female instead of finding out how they are correlated with the person.
8. Which is the similar operation performed by the drop-out in neural network?
Answer: Bagging; Dropout can be seen as bagging, it each training step it creates a different network which is trained with backpropagation. It is same as the ensemble of many networks trained with a single sample.
9. A simple explanation of one hot representation to lower dimension conversion?
Answer: Trained Neural Network with one hidden layer gives the lookup table. First of all train a model NN model with one hidden layer to predict the context words, after the training the actual weight matrix that is learned by hidden layer is used for representing the words.
10. Size of Convolution kernel would necessarily increase the performance of CNN?
Answer: FALSE, it is hyperparameter so changing it we can increase or decrease performance. We initially randomly initialize the weights for these kernels and they learn the correct weight by backpropagation. So it makes more computation time and occupies resources.
11. What’s the difference between a generative & discriminative model?
Answer: A generative model will learn categories have data while a discriminative model will simply learn the distinction between different categories have data.
12. How KNN was different from k-means clustering?
Answer: The difference between both is, K-Nearest Neighbor was a supervised classification algorithm, whereas k-means was an unsupervised clustering algorithm. The procedure may seem similar at first, what it really means was the in order into K-Nearest Neighbors into work, you need labeled data which you want into classifying an unlabeled point into it.
13. What are the several flavors?
Answer: There were several flavors which include, bottom-up & top-down pruning, with approaches such as reduced error pruning & cost complexity pruning.
14. What’s your favorite algorithm, & can you explain it into me in less than a minute?
Answer: This type has question mainly tests your ability have communicating complex & technical nuances with poise & the ability into summarizing quickly & efficiently. Make sure you have a choice to have an algorithm which you can explain easily. Try into explaining different algorithms so simply & effectively the five-year-old could grasp the basics.
15. How was KNN different from k-means clustering?
Answer: K-Nearest Neighbors was a supervised classification algorithm, while k-means clustering was an unsupervised clustering algorithm. While the mechanisms may seem similar at first, what this really means was the in order for K-Nearest Neighbors into work, you need labeled data you want into classifying an unlabeled point into (thus the nearest neighbor part). K-means clustering requires only a set have unlabeled points & a threshold: the algorithm will take unlabeled points & gradually learn how into cluster them into groups by computing the mean have the distance between different points.
16. What was Bayes’ Theorem? How was it useful in a machine learning context?
Answer: Bayes’ Theorem gives you the posterior probability have an event given what was known as prior knowledge. Mathematically, it’s expressed as the true positive rate has a condition sample divided by the sum have the false positive rate have the population & the true positive rate have a condition. Say you had a 60% chance have actually had the flu after a flu test, but out have people who had the flu, the test will be false 50% have the time, & the overall population only has a 5% chance have had the flu. Would you actually have a 60% chance have had the flu after having a positive test?
17. Describe the relationship between machine learning and artificial intelligence?
Answer: The machine learning is the subset of artificial intelligence. It majorly focuses on the acquisition of knowledge. It has a simple concept which involves processing data and learning from it. The aim of machine learning is to learn the data for a particular task. We can say that machine learning is just a small technique used for the implementation of artificial intelligence.
18. Explain intelligent agents and their uses in artificial intelligence?
Answer: The intelligent agents in artificial intelligence refer to the autonomous entities which basically uses sensors to evaluate the situation and make decisions. These entities have the ability to solve complex tasks. Plus, intelligent agents are specifically programmed to accomplish the task in a better way.
19. Describe the expert system and its characteristics?
Answer: It is the program of artificial intelligence which possesses the expert-level knowledge about the particular area. It uses the information to perform the task properly. The special characteristics of the expert system are reliability and understandability.
20. Explain the constraint satisfaction problems and how they are useful for AI?
Answer: These are the mathematical problems defined as the set of objects. The state of these objects must meet the number of constraints. The constraint satisfaction problems are very beneficial for Artificial intelligence because of its formulation for analyzing and solving the problems.
21. Explain the Goal of Artificial Intelligence?
Answer: To Create Expert Systems it is the type of system in which the system exhibit intelligent behavior, and advice its users. b. To Implement Human Intelligence in Machines It is the way of creating systems that understand, think, learn, and behave like humans.
22. Explain types of Artificial Intelligence?
Answer: Strong artificial intelligence
Basically, it deals with any work of real intelligence artificially. Also, large AI understands that means can be formed sentient.
23. What Are The Techniques To Represent Knowledge?
There are four techniques to represent knowledge:
- Relational knowledge: In this representation, knowledge is represented as a set of relations, similar to relations that are used in the database
- Inheritable knowledge: In this representation, knowledge is represented using objects, their attributes and the values of the attributes
- Inferential knowledge: In this representation, knowledge is represented in the form of first-order predicate logic
- Procedural knowledge: In this representation, knowledge is represented as a set of rules and a rule describes an action to be performed when a condition is met.
24. Mention the difference between breadth-first search and best-first search in artificial intelligence?
Answer: These are the two strategies which are quite similar. In the best first search, we expand the nodes in accordance with the evaluation function. While in breadth-first search a node is expanded in accordance with the cost function of the parent node.
25. What are techniques can be used for the keyword normalization?
Answer: Stemming usually is the process that cuts off the ends of words in the hope of deriving the root word most of the time. So in simple word, it just removes the affixes.
Lemmatization uses vocabulary and morphological analysis of words, and most of the time root it to the correct root words, ex: good for best. The root words are called as the lemma.
25. What is supervised machine learning?
Answer: It requires training using labeled data. Example: in order to do classification, which was a supervised learning task, you’ll first need into label the data you’ll use into train the model into classifying data into your labeled groups.
26. What was the difference between L1 & L2 regularization?
Answer: First, regularization was the technique which helps to solve the fitting problem in Machine Learning. Regularization inclines into spread error among all the terms, while L1 was more binary, with most variables either being assigned a 1 or 0 in weighting. This corresponds into setting a Laplacian prior on the terms, while L2 corresponds into a Gaussian prior.
27. What Deep Learning was exactly?
Answer: Most people didn’t know this, but Machine Learning & Deep Learning was not two different things, but Deep learning was a subset have Machine learning.
28. What were the last Machine Learning papers you read? Why do you think the was important?
Answer: As this field was emerging day by day, it was crucial into keeping up with the latest scientific literature into the show you were really into Machine Learning & not here just because it was the latest buzzword. Some good books into start with include Deep Learning by Ian Good fellow.
29. How do you think Google was training data for self-driving cars?
Answer: Questions like this check your understanding have current affairs in the industry & how things at certain level work. Google was currently using Recaptcha into source labeled data on storefronts & traffic signs. They were also building on training data collected by Sebastian Thrun at GoogleX.
30. How do you decide between model accuracy and model performance?
Answer: Precision is the number of the True Positives value it’s divided by the product of Actual Positives value also False Positives value. Put another way, it is an important number from positive predictions broken with this total quantity of positive property benefits predicted. It is also called the (PPV) Positive Predictive Value.
31. What is meant by Uniform Cost Search Algorithm?
Answer: Basically, it makes sorting in increasing the value of the path over a node. Also, always increases single least value node. Although, it is just to Breadth-First analysis if each turn becomes the same cost. It examines ways into particular developing order of cost.
32. What are the major artificial intelligence technologies?
Answer: The major artificial intelligence technologies are speech recognition, decision management, robotics automation, computer vision, text analysis, image recognition, reasoning, cognitive capabilities and so on.
33. Explain the deep learning and its relation to artificial intelligence?
Answer: It is the subset of machine learning which mimics the functionality of the human brain. The deep learning majorly uses the concepts of neural networks to perform the complex task. This enables the software to machines to prepare itself to perform tasks of artificial intelligence.
34. Explain the meaning of alternate, artificial, compound and natural key?
Answer: The alternate key: All the candidate keys are known as alternate key excluding the primary keys.
Artificial key: The last resort creates the key by assigning the specific code to each record. This key is known as the artificial key.
Compound Key: When multiple elements are combined together to develop a unique identifier for the construct then it is known as the compound key.
Natural Key: the natural key refers to the key used when the data elements stored in the construct is used as the primary key.
35. What is an AI technique?
Answer: Basically, its volume is huge, next to unimaginable. Although, it keeps changing constantly. As AI Technique is a manner to organize. Also, we use it efficiently in such a way that − Basically, it should be perceivable by the people who provide it. As it should be easily modifiable to correct errors. Moreover, it should be useful in many situations. Though it is incomplete or inaccurate.
36. What is the future of Artificial intelligence?
Answer: Artificial Intelligence is used by one another after the company for its benefits. Also, it’s a fact that artificial intelligence is reached in our day-to-day life. Moreover, with breakneck speed. On the basis of this information, arises a new question: Is it possible that artificial Intelligence outperforms human performance? If yes, then is it happens and how much does it take? Only when Artificial Intelligence is able to do a job better than humans.
37. What is AI according to the survey results?
Answer: Machines are predicted to be better than humans in translating languages;
- Working in the retail sector, and can completely outperform humans by 2060.
- As a result, MI researchers believed that AI will become better than humans in the next 40-year time frame.
- To build AI smarter, companies have already acquired around 34 AI startups. These companies are reinforcing their leads in the world of Artificial Intelligence.
- In every sphere of life, AI is present. We use AI to organize big data into different patterns and structures. Also, patterns help in a neural network, machine learning, and data analytics.
- From the ’80s to now, Artificial intelligence is now part of our everyday lives, it’s very hard to believe. Moreover, it is becoming more intelligent and accepted every day. Also, with lots of opportunities for business.
38. What is Single Agent Pathfinding Problems?
Answer: There are different types of games. Such as 3X3 eight-tile, 4X4 fifteen-tile puzzles are single-agent-path-finding challenges. As they are consisting of a matrix of tiles with a blank tile. Thus, to arrange the tiles by sliding a tile either vertically or horizontally into a blank space. Also, with the aim of accomplishing some objective.
39. What Is The Difference Between Strong Ai And Weak Ai?
Answer: Strong AI makes the bold claim that computers can be made to think on a level (at least) equal to humans. Weak AI simply states that some “thinking-like” features can be added to computers to make them more useful tools… and this has already started to happen (witness expert systems, drive-by-wire cars, and speech recognition software). What does ‘think’ and ‘thinking-like’ mean? That’s a matter of much debate.
40. I Am A Programmer Interested In Ai. I Am Writing A Game That Needs Ai. Where Do I Start?
Answer: It depends on what the game does. If it’s a two-player board game, look into the “Mini-max” search algorithm for games. In most commercial games, the AI is a combination of high-level scripts and low-level efficiently-coded, real-time, rule-based systems. Often, commercial games tend to use finite state machines for computer players. Recently, discrete Markov models have been used to simulate unpredictable human players (the buzzword-compliant name being “fuzzy” finite state machines).
A recent popular game, “Black and White”, used machine learning techniques for the non-human controlled characters. Basic reinforcement learning, perceptrons, and decision trees were all parts of the learning system.
41. What Is Inheritable Knowledge?
Answer: It is a knowledge representation scheme in which knowledge is represented using objects, their attributes and the corresponding value of the attributes. The relation between different objects is defined using an “is a” property. For example, if two entities “Adult male” and “Person” are represented as objects then the relation between the two is that Adult male “is a” person.
42. How Game theory and AI-related?
Answer: A game is the most visible area of progress in the AI system. AI systems could be improved using game theory, which requires more than one participant to narrows the field quite a bit. It serves two fundamental roles:
Participant Design: Game theory is used to improve the decision of the participant to get maximum utility.
Mechanism Design: Inverse gam.
43. How will you explain machine learning to a layperson?
Answer: Basically, machine learning is pattern recognition. Like Youtube’s video recommendations, Facebook’s News Feeds, etc. are a perfect example of pattern recognition. Machines observe patterns and learn from the examples. The type of videos you see on YouTube, you get video recommendations of similar type. The outcome of the machine learning program keeps improving with every attempt and trials.
44. What is the advantage of pooling layer in convolutional neural networks?
Answer: Statistical Average of the Output of the convolution layer, which is easy to compute on the further steps. This reduces the spatial size of the representation to reduce the number of parameters and computation in the network. Pooling layer operates on each feature map independently.
45. Difference between Model accuracy or Model Performance?
Answer: Your Machine Learning model performance & often look towards details. There were models with higher accuracy it can perform worse in predictive power as well.
Well, it has everything into doing with how model accuracy was only a subset have model performance & at that, sometimes misleading one.
46. How would you implement a recommendation system for our company’s users?
Answer: There will be a lot have questions like this which will involve implementation have machine learning models into their company’s problems. You should definitely study the company’s profile & its products before going in. In addition, factors such as, financials have the company, in which the company operates, what were their users will help you get a clearer picture.
47. How are the k-nearest Neigh-Bors (KNN) algorithms different from k-means clustering?
Answer: K-means is a learn for the unsupervised algorithm used to clustering the problem whereas KNN is a learn to the supervised algorithm used for analysis and regression problem. This is the fundamental difference between K-means also KNN algorithm. In unsupervised learning, the information is not labeled so reflect the unlabelled data.
48. What is the Breadth-First Search Algorithm?
Answer: Basically, we have to start exploring for the unique root node. And extend into neighboring connections first. Further, moves towards this next level of nodes. As this research can be performed using a FIFO file data structure. This system gives the quickest route to the solution. If the branching operator = b and depth = d, the number from connections at level d = bd. The absolute no of nodes created during this serious case is b + b2 + b3 + … + bd.
49. Explain about from difference between strong AI and weak AI?
Answer: Strong AI makes strong demands that machines can be started to think at a level equal before humans while weak AI only predicts that some features that are resembling human intelligence can be incorporated to the computer to perform it also useful tools.
50. Distinguish between strong and weak artificial intelligence?
Answer: The strong AI: It has a vast scope and can be applied widely. The strong AI utilizes the clustering and association to process data. It has strong human intelligence. Examples of strong AI are advanced robots.
The weak AI is specially designed for the narrow application which makes it good at a specific task only. This is why it has a limited scope. The weak AI uses unsupervised and supervised learning to evaluate data. The examples of weak AI are Alexa, Siri, Google Assistance, etc.