User:Waltz2u/sandbox

The International Society of Data Scientists Inc, or ISODS for short, registered as a Massachusetts charitable non-profit, is a professional organization for Data Science and AI practitioners and researchers worldwide, including Data Scientists, Machine Learning Scientists, AI Scientists, Data Analysts, Data Engineers, Software Engineers, Risk Analysts, Actuaries, Business Analysts, etc., who apply Data Science and AI at work.

ISODS promotes Data Science and AI domestically in the United States as well as internationally via activities such as competitions, conferences, professional exams, and publications.

Membership
Members are categorized as individual and affiliate members, both need to be registered with the Society. Individual members pay due; while affiliate members do not pay due. Both individual and affiliate members may take professional exams toward 3 titles: Associate Master, Master, and Grandmaster.

Competitions
In 2022, the Society hosts the 3rd Annual International Competition in Data Science & AI with the participation of students internationally in 3 tracks based on Kaggle, including Natural Language Processing, Computer Vision, and Structured Data, and a Statistical Data Science track separately. The 2nd Annual International Competition was in 2020-2021, and the inaugural competition was in 2019.

Probability (P) (multiple-choice exam)
Basic probability concepts; Discrete and continuous univariate random variables (including binomial, negative binomial, geometric, hypergeometric, Poisson, uniform, exponential, gamma, normal, and mixed) and their applications. Multivariate random variables and their applications.

Statistics (STAT) (multiple-choice exam)
Discrete and continuous random variables, exponential family, joint and marginal and conditional distributions, order statistics, statistical inference: point estimation, confidence interval estimation, and hypothesis testing, the central limit theorem, sums of random variables, independence

Linear Algebra (LA) (multiple-choice exam)
Introduction to linear algebra with elementary applications. covers the following major topics: linear systems of equations, matrices, determinants, linear transformations, eigenvalues and eigenvectors.

Calculus (CAL) (multiple-choice exam)
Introduction to differential and integral calculus. The main topics it covers are limits, derivatives, integrals, the Fundamental Theorem of Calculus, and some basic applications of these ideas. Transcendental functions, formal integration, polar coordinates, infinite sequences and series, parametric equations.

Predictive Analytics (PA) (project-based)
Modeling language: RStudio/R or Python equivalence. Problem Analysis, Data Visualization, Data Types and Exploration, Data Issues and Resolutions; Generalized Linear Models, Decision Trees, Clustering and Principal Component Analysis.

Database Management (DM) (project-based)
Structured data: Basic database design and implementation concepts. Database design techniques, including relational design and E-R analysis. Database queries using SQL. Unstructured data: nonSQL. Semi-structured data.

Machine Learning (ML) (multiple-choice/essay exam)
Classification and Regression. Logistic Regression, Decision Trees, Ensembles, Neural Networks, SVM, Naïve Bayes, KNN, Clustering, Recommendation Systems

Big Data Analytics (BDA) (project-based)
Modeling language: Python/Java using Spark/Hadoop. RDD, DataFrames, and SparkSQL. Typical classification and regression models in the Big Data world.

Object-oriented Programming (PRG) (problems)
Modeling language: Python/Java. Object-oriented Concepts such as classes, objects, data abstraction, methods, method overloading, encapsulation, inheritance and polymorphism.

Data Structures and Algorithms (DSA) (problems)
Modeling language: Python/Java. Arrays, Strings, Stacks, Queues, Linked Lists, Hash Tables, Trees, and Graphs. Asymptotic analysis (Big-O notation). Recursion and Dynamic Programming. Greedy Algorithms. Divide and Conquer. Backtracking. Master Theorem. Sorting and Searching.

Time Series (TS) (project-based)
Modeling language: RStudio/R or Python equivalence. Time series regression and exploratory data analysis, ARMA/ARIMA models, model identification/estimation/linear operators, Fourier analysis, spectral estimation, and state space models Preprocessing: smoothing data, estimate peaks, distribution, ...

Deep Learning 1 (DL1) (project-based)
Modeling language: Python with relevant packages. MLP foundation: layers, activations, loss functions, underfitting/overfitting, model selection, forward/backward propagation, dropout, etc.. Convolutional Neural Networks foundation: Convolution, padding, stride, pooling, etc. Multiple input/output channels. Classic and modern architectures: LeNet, AlexNet, VGG, NiN, GoogleLeNet, ResNet, DenseNet, Xception, EfficientNet, etc. Reading, writing, and visualizing image data. Image Augmentation. Transfer Learning. Style Transfer. Optimization Algorithms (SGD, RMSProp, etc.). Object detection: Bounding boxes. Anchor boxes.

Deep Learning 2 (DL2) (project-based)
Modeling language: Python with relevant packages. Recurrent Neural Networks foundations. Text processing. Gated Recurrent Units (GRU). Long Short-Term Memory (LSTM). Deep Recurrent Neural Networks. Encoder-Decoder Architecture. Sequence to Sequence Learning. Beam Search. Attention Mechanisms. Word Embedding. Pretraining.

Reinforcement Learning (RL) (project-based)
Modeling language: Python with relevant packages. Markov decision processes, value functions, Monte Carlo estimation, dynamic programming, temporal difference learning, eligibility traces, and function approximation

Publications
The Journal of Data Science and Artificial Intelligence publishes the first issue in 2022

Conferences
ISODS organizes conferences in Applied Data Science & Artificial Intelligence