AI assistants are generally the application programming interfaces (APIs) or automated machines/gadgets that are designed to perform certain “non-value adding services” like task-scheduling, setting and/or executing reminders, email management, etc. https//hcitexpert.com/2018/02/glossary-of-ter

It is a set of protocols or subroutine definitions for building an application software computing interface. APIs are designed to meet the objective of the end-users by allowing the devices to share data and enhancing interoperability and functionality between applications, devices, and individuals. The functionalities of the API are monitoring and controlling the interactions between multiple software intermediaries through calls or requests made, using required data formats, following suitable conventions, etc. An example of API is whether snippets are shared through smartphones and PCs or the Google web page. Although none of the above are associated with weather forecasting as a business. Another example is to log in to an app or any website using Google or Facebook or Twitter or Github account details.

It is a broad field of computer science and information technology that makes these systems capable of performing tasks that typically require human intelligence. Thus, a machine’s capability in decision-making vis-à-vis simulating human intelligence and behavior to address any task or situation is called AI.

A computerized “learning model” that acts like the brain of a human being that can solve tasks that are complicated and tough for traditional computer systems to solve.

It is the short form of “CHAT roBOT”. It is designed to simulate the conversation and communication style of a human being (with human users) through text chats, voice commands, or both. Chatbots make use of artificial intelligence to act as a commonly used interface for computer programs.

It is a section of machine learning (ML) that mimics the neural network of the human brain to filter data through self-adjusting virtual networks of neurons. Thus, the computer system develops cascading layers of the virtual neuronal network to mimic human thought patterns through artificial neural networks.

It’s used in a broad sense that incorporates machine learning, deep learning, and classical learning algorithms.

can be defined as an application vis-à-vis a subset of AI that provides a machine the ability to learn automatically and improve from experience that it has been explicitly programmed for through systems or software models. AI evolved in the 1950s while ML evolved as a subset of AI in the 1980s and deep-learning evolved in 2010 as a subset of ML. Example Google duplex, an application of Google Assistant uses AI to take some real-world tasks such as booking an appointment.

It is a kind of computer algorithm that receives human language as input and then converts it into machine-understandable representations. A natural language is also called ordinary language. It is defined as a language that has evolved as a spoken mode for communication among humans without conscious planning or premeditation. Natural language processing (NLP) is a pool that integrates concepts and expertise from diverse fields, viz. information engineering, linguistics, computer science, and artificial intelligence. The broad areas of NLP involve speech recognition, natural language understanding (NLU), and natural language generation (NLG) (https //en.wikipedia.org/wiki/Natural_language_processing).

A type of neural network that makes use of the outputs of the previous step (or layer) to sense the sequential information and then make use of it as input for the present layer. Thus, it recognizes patterns, and creates outputs based on those calculations.

Natural Language Processing (NLP) is the Algorithm that processes human language input and converts it into understandable representations. A natural language is also called ordinary language. It is defined as a language that has evolved as a spoken mode for communication among humans without conscious planning or premeditation. Natural language processing (NLP) is an amalgamation of several diverse fields, viz. linguistics, computer science, information engineering, and artificial intelligence. Challenges in natural language processing frequently involve speech recognition, natural language understanding, and natural language generation (https //en.wikipedia.org/wiki/Natural_language_processing) .

A type of neural network that makes sense of sequential information and recognizes patterns, and creates outputs based on those calculations.

Software that experiments with different actions to figure out how to maximize a virtual reward, such as scoring points in a game.

Robotics is the interdisciplinary domain of electronics, computer science, mechanical, and other relevant branches of engineering that designs and develops robots. The internal machinery of computer systems receives the sensory feedback, and information processing, thereof. The sensory organs of a robot capture images, sounds, ambient conditions, etc through computerized sensors.

It refers to showing a learning software labeled example data (viz. photographs or diagrams or structures or models), to edify a computer on what to do. Supervised learning is a machine learning technique that uses labeled datasets to train the models to generate the desired algorithms. The prediction model is then used to predict the output for future or unseen data. Training data are needed to be fed to the learning algorithm to discover patterns in the training data to generate an ML model that learns from the algorithm.

It is a learning algorithm used in machine learning without annotated examples. Inferences are drawn by the algorithm from the unlabeled input datasets. Example a cluster analysis.

Virtual assistants are electronic gadgets that serve as an assistant by recognizing and analyzing the voice and/or text inputs through natural language processing (NLP) and thereafter translating the same into executable commands. These assistants make the learning process continues and improves accuracy through repeated training and testing using AI/ML. Examples Apple’s Siri, Amazon’s Alexa, and Google Now.

It is measured as the ratio between the “total number of correct predictions made” to the “total number of predictions”, which is represented as (TP+TN)/(TP+TN+FP+FN). Accuracy is better to use when the target variable data in different classes is nearly balanced. Accuracy should never be used as a measure of performance when the majority of the target variable data belongs to a single class.

It is a set of instructions that dictates the rule to a computer system, to train artificial intelligence or neural network, to enable it to learn the patterns or make predictions on its own; Examples of popular algorithms are: clustering, recommendation, classification, etc.

It’s a Python package (available for Python 2.7 and Python 3) to parse HTML and extended mark-up language (XML) documents (even that with malformed markup, i.e. non-closed tags, so named after tag soup) by generating a parse tree for the parsed online HTML/XML pages to extract data for web scraping.

Big data is a “field” that is associated with scientific methods to analyze, systematically extract information from, or otherwise deal with data sets that are too large or complex to be dealt with by traditional data-processing application software. Data with many cases (rows) offer greater statistical power, while data with higher complexity (more attributes or columns) may lead to a higher false discovery rate. Big data challenges include capturing data, data storage, data analysis, search, sharing, transfer, visualization, querying, updating, information privacy, and data source. The key concept of big data was originally summarized by the 3 V’s: volume, variety, and velocity.

Classification algorithms are the techniques of supervised learning that enable machines to identify the category of new observations based on the training data. Various classification algorithms are support vector machines, logistic regression, naive Bayes, decision trees, and K-nearest neighbors.

It’s a kind of unsupervised learning in machine learning aimed at automatically discovering natural grouping in data. Clustering is used for exploratory data analysis to discover hidden or unidentified patterns and also the grouping of unlabelled data. Clustering algorithms principally work on different methods, like, distribution-based, density based, centroid-based, and hierarchical-based. Different popular clustering algorithms are the K-means clustering algorithm, Balance Iterative Reducing and Clustering using Hierarchies (BIRCH) algorithm, Gaussian Mixture Model algorithm, density-based spatial clustering of applications with noise (DBSCAN), Ordering Points to Identify the Clustering Structure (OPTICS), Affinity Propagation clustering algorithm, Mean-Shift clustering algorithm, and Agglomerative Hierarchy clustering algorithm.

Same as Cluster Analysis algorithm. Please see above

It is a computerized simulation of the thinking pattern of human brains through self-learning by data mining, pattern recognition, natural language processing (NLP), etc.

CNN (also abbreviated as “ConvNet”) is a type of neural network, used in deep learning, that identifies and makes sense of images. CNN's are characterized by their shared-weights architecture and feature of translational invariance. Hence, CNNs are termed shift-invariant artificial neural networks or space invariant ANNs (SIANN). CNN is a deep neural network used to analyze images including medical images, image classification, and image and video recognition. CNNs are also used even in natural language processing, recommender systems, and financial time series.

• CNNs are advanced “multilayer perceptrons” featured by “fully connected networks of neurons of multiple connected layers”.
• The flaw of the "full connectedness" of the neuronal networks is the overfitting of data during modeling.

Data mining is defined as the advanced method of analyzing large datasets to discover known and unknown patterns. Pattern recognition involves methods and applications of statistics, database systems, and machine learning to discover patterns in datasets. Data mining makes use of knowledge and skills of computer science vis-à-vis advanced statistics to extract information from a data set using artificial intelligence and transform the information into a comprehensible structure for further use. Data mining makes use of "knowledge discovery in databases" (KDD) through analyzing data sets to discover patterns hidden in the data.

It’s an interdisciplinary field consisting of statistical systems and processes, and that of information science, and computer science to provide insight from structured or unstructured data. Data science amalgamates various scientific techniques of data structuring and management, cleaning, analysis, and extracting meaningful information by scientific processes, algorithms, and systems.

Data wrangling or data munging is a method of manually converting and mapping data from a raw format to another format to make it more valuable and acceptable for advanced tasks viz. data analytics and machine learning. It includes munging and data visualization. A data wrangler is a person who performs various transformation operations on data, including data-munging, data aggregation, data visualization, training of the statistical model, etc.

Decision tree methods are a type of “supervised machine learning” algorithm where a tree and branch-based model are used to map decisions through the continuous splitting of data based on certain parameter(s). The ID3 algorithm (developed by Quinlan) which uses a top-down, greedy approach, is used for the decision tree method.

It is designed to analyze inputs like images or audio through virtual neuronal layers of which the previous layer serves as input for the next layer and ultimately an output or activation function.

It is a single score that represents precision and recall = (2*Precision*Recall)/(Precision+Recall). F1 score is the harmonic mean (HM) of Precision & Recall. A harmonic mean is used when the sample data contains extreme values because it is more balanced than arithmetic mean.

an open-source Python library used for analyzing multi-dimensional, large arrays and matrices. It was initially released in 1995 as “Numeric” and later in the name of “NumPy” in 2006. This library maintains a huge compilation of high-level mathematical functions for matrix/ array operations.

• The library “Numeric” was the ancestor of NumPy, developed by Jim Hugunin and several other developers. NumPy was created by Travis Oliphant in 2005 integrating “Numarray” into Numeric and several other modifications. NumPy is open-source software and has many contributors.

It is an open-source library (under the three-clause BSD license) written for Python programming for data manipulation and analysis. It was initially released in 2008. Pandas library includes the data structures and operations needed for manipulating numerical tables and time series. The name Pandas have been derived from "panel data" which is an econometrics term for data sets kept under observations over multiple periods.

Closeness of two or more measurements with each other= TP/(TP+FP).

It’s a statistical technique to extract information from data and thereby use it to predict trends and behavior patterns–for example, sales forecasts, predicting customer churn, and industrial machine failure.

measures the proportion of actual positives that are correct identified= TP/(TP+FN). Thus, precision is about being precise, and recall is about capturing all the cases. Precision gives us information about the classifier’s performance concerning false positives. Recall gives us information about the classifier’s performance concerning false negatives. When we want to minimize false negatives, we will like to have 100% recall, but with good precision. Conversely, when we want to minimize false positives, we will like to have 100% precision.

It is a freely available machine learning library used in Python programming. Scikit Learn was formerly known as scikits learn and is abbreviated as sklearn. Scikit Learn is a powerful package that can be used for classification, regression, and clustering algorithms (random forests, k-means, DBSCAN, gradient boosting, support vector machines). Besides, it can interoperate with numerical and scientific libraries: NumPy and SciPy, respectively. Scikit-learn integrates well with many other Python libraries, such as matplotlib and plotly for plotting, numpy for array vectorization, pandas dataframes, scipy, and many more.

this open-source library which has been built on NumPy stack, is used for scientific computing using Python. The important modules of SciPy are linear algebra, data optimization integration, interpolation, FFT, signal and image processing, etc. It was initially released in 2001.

Seaborn is a library built on top of matplotlib and integrated with PyData stack for creating attractive and informative statistical graphs. It also supports numpy and pandas statistical routines and data structures. It internally performs statistical aggregation and semantic mapping to produce informative plots. This makes it different from other visualization libraries.

Proportion of actual negatives that are identified correctly= TN/(TN+FP). Specificity is just the converse of recall.