AI (Artificial Intelligence)
AI refers to the simulation of human intelligence in machines. It is the theory and development of computer systems able to perform tasks that typically require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.
API (Application Programming Interface)
API is a set of tools and protocols that allows different software applications to communicate with each other. It defines the kinds of requests that can be made, how to make them, the data formats used, etc.
Activation Function
Activation functions determine the output of a neural network, its accuracy, and the computational efficiency of training a model. They play a crucial role in regulating the information that should be passed to the next layer.
Algorithm
An algorithm is a set of unambiguous instructions that, when followed, solves a problem or performs a task. In the context of AI/ML, algorithms are used to find solutions or make decisions based on data.
Algorithmic Bias
Algorithmic Bias refers to the presence of non-objective, unfair, and often harmful biases in computer algorithms due to flawed assumptions or methodologies during the algorithm development process.
Anomaly Detection
Anomaly detection refers to identifying abnormal or rare items in a data set that do not conform to general patterns. It’s crucial for identifying fraudulent activities, system errors, and outliers in statistical studies.
Augmented Reality (AR)
AR integrates digital information with the user’s environment in real-time. Unlike virtual reality, which creates a totally artificial environment, augmented reality uses the existing environment and overlays new information on top of it.
Autoencoder
An autoencoder is a type of neural network used to learn efficient codings of input data, typically for the purpose of dimensionality reduction or feature learning.
Back-End Development
Back-end development refers to server-side development, the backbone of a web application, where the logical operations happen. It interacts between the database and the browser (front-end), sending data to be displayed.
Backpropagation
Backpropagation is a supervised learning algorithm for training multi-layer perceptrons (Artificial Neural Networks). It calculates the gradient of the loss function with respect to its parameters, optimizing the weights to minimize error.
Bayesian Network
A Bayesian Network is a probabilistic graphical model that represents a set of variables and their conditional dependencies via a directed acyclic graph.
Bias
In AI/ML, bias refers to errors and inaccuracies in a model’s predictions due to flawed assumptions, representation, or design in the algorithm, often leading to unfair or unethical outcomes.
Black Box Model
Black Box Model is a system where the internal workings are not understood or accessible, and only the inputs and outputs are known. In AI, this often refers to complex models where the decision-making process is not transparent or explainable.
Blockchain
Blockchain is a decentralized and distributed ledger technology that securely records information across several computers to ensure that it can’t be changed retroactively, providing a high level of data integrity.
CNN (Convolutional Neural Network)
CNN is a deep learning algorithm which can take in an input image, assign importance to various aspects/objects in the image, and be able to differentiate one from the other. Predominantly used in image and video recognition.
CUDA
CUDA is a parallel computing platform and programming model developed by Nvidia for general computing on its own GPUs (graphics processing units), enabling dramatic increases in computing performance.
Chatbot
A chatbot is a software application designed to simulate human conversation. It can interact with users using textual or auditory methods and is used for customer service, information acquisition, or other user interactions.
Classification
Classification, in the context of machine learning, is a type of supervised learning where the goal is to predict the categorical class labels of new instances, based on past observations.
Cloud Computing
Cloud Computing allows for scalable AI computations by providing virtualized computing resources over the internet, offering flexibility in storage and computational power.
Clustering
Clustering is a type of unsupervised learning that automatically groups similar objects into sets or ‘clusters’. It is used to analyze raw data and form structured groups.
Computer Vision
Computer vision enables computers to interpret and make decisions based on visual data, with applications in facial recognition, object detection, autonomous vehicles, and many others.
Cross-Validation
Cross-validation is a resampling procedure used to evaluate machine learning models on a limited data sample, with the goal of reducing overfitting and providing insight into how the model will generalize to an independent dataset.
Cryptography
Cryptography is the study of secure communication techniques, ensuring data privacy, integrity, and authentication in digital systems, including AI.
Cybersecurity
Cybersecurity is the practice of protecting systems, networks, and programs from digital attacks, particularly important in AI implementations where data integrity and privacy are paramount.
Data Augmentation
Data Augmentation involves creating new training samples by applying various transformations to the existing dataset, enhancing the model’s performance and generalization capabilities by exposing it to more diverse data points.
Data Lake
Data Lake is a storage system that holds a vast amount of raw data in its native format until it’s needed, serving as a single repository, beneficial for AI systems that require large and diverse datasets.
Data Privacy
Data Privacy concerns the handling, processing, storage, and protection of personal information. It is paramount in AI, given the vast amounts of sensitive data AI systems often interact with.
Decision Trees
Decision Trees are a type of Supervised Learning Algorithm that is mostly used for classification problems. It works for both categorical and continuous input and output variables.
Decryption
Decryption is the reverse process of encryption, converting encoded data back into its original form, allowing for the secure transmission of data in AI systems.
Digital Twin
A Digital Twin is a digital replica of physical assets, processes, or systems. It provides a dynamic model that simulates the physical counterpart. AI analyzes the data to extract insights and optimize the operation.
Dimensionality Reduction
Dimensionality Reduction is the process of reducing the number of random variables under consideration by obtaining a set of principal variables, essential for improving model efficiency and dealing with the “curse of dimensionality.”
Edge Computing
Edge Computing involves data processing at the edge of the network, closer to the data source, reducing latency and bandwidth use, ideal for real-time AI applications.
Encryption
Encryption is the process of converting data or information into code to prevent unauthorized access, playing a critical role in ensuring the security of AI systems.
Ensemble Learning
Ensemble Learning methods use multiple learning algorithms to obtain better predictive performance than could be obtained from any of the constituent learning algorithms alone.
Feature Engineering
Feature Engineering is the process of using domain knowledge to create features (feature vectors) that make machine learning algorithms work more efficiently and effectively, often enhancing model performance.
Front-End Development
Front-End Development deals with the user interface and user experience of a website or web application, utilizing technologies like HTML, CSS, and JavaScript, essential for AI-driven websites to ensure user engagement.
Fuzzy Logic
Fuzzy Logic is a computing approach based on “degrees of truth” rather than the usual “true or false” (1 or 0) binary logic, allowing for modeling complex systems with an imprecise spectrum of values.
GAN (Generative Adversarial Network)
GANs consist of two networks, a generator and a discriminator. The generator creates images from noise, while the discriminator evaluates them. They are used to generate realistic images, sound, and text.
Genetic Algorithm
Genetic algorithms are a part of evolutionary algorithms used to solve search and optimization problems. They are based on the process of natural selection and combine inheritance, mutation, selection, and crossover to arrive at a solution.
Gradient Descent
Gradient Descent is an optimization algorithm used to minimize the errors in predictions by iteratively moving towards the minimum of the cost function, usually used in training machine learning models.
Grid Search
Grid Search is a hyperparameter optimization technique used to find the optimal hyperparameters for a model, involving a systematic exploration of all possible combinations of the hyperparameter values.
Hyperparameter
Hyperparameters are external configurations for algorithms which are not learned from the data. They are set prior to the commencement of the learning process, influencing the speed and quality of the learning process.
Hyperparameter Tuning
Hyperparameter Tuning refers to the process of adjusting the configuration settings used to structure machine learning models to improve their performance.
Image Recognition
Image recognition refers to the ability of software to identify objects, places, people, writing, and actions in images. It is used in various applications like biometrics, healthcare, and retail to recognize and analyze features.
K-Means
K-Means is a clustering algorithm that divides a set of data points into ‘K’ number of mutually exclusive clusters. It is popular for cluster analysis in data mining and machine learning.
Knowledge Graph
A Knowledge Graph represents a collection of interlinked descriptions of entities – real-world objects, events, situations, or abstract concepts – with free-form semantics.
LDA (Linear Discriminant Analysis)
LDA is a method used in statistics, pattern recognition, and machine learning to find a linear combination of features that best separates two or more classes of objects or events.
Latent Variable
Latent variables, in statistical modeling, are variables that are not directly observed but are rather inferred through a mathematical model from other variables that are observed.
Linear Regression
Linear Regression is a statistical approach for modelling the relationship between a dependent variable and one or more independent variables.
Load Balancing
Load Balancing distributes incoming network traffic or computational tasks across various servers, ensuring no single server is overwhelmed, crucial for AI systems that require real-time processing on large datasets.
Logistic Regression
Logistic Regression is a statistical method for analyzing a dataset in which there are one or more independent variables that determine an outcome represented by a binary variable, for example, yes/no or 0/1.
Loss Function
A Loss Function measures the difference between the model’s prediction and the true label. It is used during the training of a model to measure errors and optimize the model.
Meta-Learning
Meta-Learning is a subfield of machine learning where automatic learning algorithms are applied to meta-data about machine learning experiments. It aims to make machine learning systems more automated and efficient.
Mixed Reality (MR)
Mixed Reality combines both VR and AR, allowing for the interaction of physical and digital objects in real-time, often enhanced with AI algorithms for more realistic interactions.
Naïve Bayes
Naïve Bayes classifiers are a collection of classification algorithms based on Bayes’ Theorem, assuming an independence between predictors, making them suitable for high-dimensional datasets.
Neural Architecture Search
Neural Architecture Search (NAS) refers to the automated design of neural networks, an approach to finding the best-performing model architecture for a given dataset and task.
Normalization
Normalization is a scaling technique where the values are shifted and rescaled so they end up ranging between 0 and 1, often used in machine learning to improve the performance and training stability of models.
One-Hot Encoding
One-Hot Encoding is a representation method for categorical data, converting each category value into a new categorical column and assigns a binary value of 1 or 0. It’s crucial for machine learning algorithms that require numerical input.
Outlier Detection
Outlier Detection is similar to anomaly detection and involves identifying the rare items in a dataset that significantly differ from the majority, which can have substantial impacts on data analysis and model training.
Overfitting
Overfitting occurs when a model learns the training data too well, capturing noise in the training data as if it were a real pattern, leading to poor generalization to new data.
PCA (Principal Component Analysis)
PCA is a dimensionality reduction technique that transforms high-dimensional data into a lower-dimensional form, preserving as much variance as possible, commonly used in exploratory data analysis and predictive modeling.
Precision
In machine learning, precision is a measure of the accuracy of a classification model, calculated as the number of true positives divided by the number of true positives plus the number of false positives.
Quantum Computing
Quantum Computing leverages the principles of quantum mechanics to process information in ways that traditional computers can’t. It holds the potential to solve complex problems much more efficiently.
RNN (Recurrent Neural Network)
RNNs are a class of neural networks that are effective in modeling sequence data. They are designed to recognize patterns in sequences of data, such as text, genomes, handwriting, the spoken word, and numerical times series data.
Random Forest
Random Forest is an ensemble learning method, constructing a multitude of decision trees at training time and outputting the class that is the mode of the classes or mean prediction of the individual trees for regression tasks.
Regularization
Regularization is a technique used to prevent overfitting in machine learning models by adding a penalty term to the loss function, constraining the complexity of the model.
Robotics
Robotics is a branch of technology that deals with the design, construction, operation, and use of robots, enabling the development of machines that can substitute for humans and replicate human actions.
SDK (Software Development Kit)
A Software Development Kit is a collection of software development tools and libraries designed to assist developers in creating applications for a certain software package or hardware platform.
SVM (Support Vector Machine)
Support Vector Machines are supervised learning models, useful for classification and regression analysis. They are effective in high-dimensional spaces and are best suited for situations where the classes are linearly separable.
Scalability
Scalability in AI systems refers to the capability of a system to handle a growing amount of work, its potential to be enlarged in order to accommodate that growth.
Semantic Analysis
Semantic Analysis is the process of relating syntactic structures, from the levels of phrases, clauses, sentences, and paragraphs to the level of the writing as a whole, to their language-independent meanings.
Sentiment Analysis
Sentiment Analysis is the use of natural language processing to identify, extract, quantify, and study affective states and subjective information, usually from source materials such as social media and customer reviews.
Sequence-to-Sequence Model
Sequence-to-Sequence Models are neural network models often used in tasks like language translation or chatbot dialogue systems where the input and output lengths can vary.
Serverless Computing
Serverless Computing lets developers build and run applications without thinking about servers, automatically managing the infrastructure needed for AI applications.
Standardization
Standardization involves rescaling the features so that they have the properties of a standard normal distribution with a mean of zero and a standard deviation of one.
Swarm Intelligence
Swarm Intelligence refers to the collective behavior of decentralized, self-organized systems, natural or artificial, often applied in optimization, routing, and clustering problems inspired by behaviors of colonies of biological entities.
TPU (Tensor Processing Unit)
A Tensor Processing Unit (TPU) is a type of application-specific integrated circuit developed by Google specifically for neural network machine learning, offering high throughput for low-precision arithmetic.
Text Mining
Text Mining involves extracting valuable information from text data. This process allows for pattern recognition, sentiment analysis, and other analyses to extract insights from unstructured data sources, like websites and social media.
Time Series Analysis
Time Series Analysis comprises methods for analyzing time series data to extract meaningful statistics and identify characteristics. It is pivotal in forecasting, signal processing, and anomaly detection in sequential data.
Transfer Learning
Transfer Learning is a machine learning method where a model developed for a task is reused as the starting point for a model on a second task, allowing for the leveraging of pre-learned knowledge on new, similar tasks.
Underfitting
Underfitting occurs when a model cannot capture the underlying trend of the data. Typically, it means the model is too simple to handle the complexity of the data, resulting in poor performance.
Validation Set
A Validation Set is a set of examples used to tune the parameters of a classifier, i.e. to avoid overfitting, in the context of machine learning. It provides a check against overfitting during the training of the model.
Virtual Reality (VR)
Virtual Reality is an interactive, computer-generated simulation of a 3D environment, offering immersive experiences and often integrated with AI to enhance interactivity.
Web Scraping
Web Scraping is the process of extracting information from websites. It is crucial for gathering large datasets required for training and testing AI models, especially in NLP and ML applications.
Weight
In neural networks, weights are used to apply leverage to the input data and are pivotal in adjusting the output of the model during the training phase. They are the learned parameters that the network uses to make predictions.
White Box Model
A White Box Model in AI is a system where the internal workings are understood and can be viewed and examined, promoting transparency, easier debugging, and trust.
Word Embedding
Word Embedding is a technique in natural language processing where words or phrases from vocabulary are mapped to vectors of real numbers. It is essential for capturing context and semantic meanings in large texts.
XAI (Explainable AI)
Explainable AI refers to methods and techniques in the application of artificial intelligence technology (AI) such that the results of the solution can be understood by human experts.
Zero-Shot Learning
Zero-Shot Learning is a type of learning where the model is able to recognise objects that it has never seen during training. It’s particularly useful when the availability of annotated data is scarce.
AI Tech Connection 