Using Generative AI for Predecting Market Movement

Meet the instructor, a seasoned finance professional turned data scientist and algo trader, who blends risk analysis, data science, and GPT-driven finance use cases.

Set up your computer by downloading Anaconda Navigator, an aggregate of applications that enhances programming across Python and other languages, then run your first line of code.

Learn to install and configure Anaconda for Python and algorithmic trading, including choosing the right operating system installer, adding Anaconda to your path, and starting with Anaconda3.

Learn the basics of Python commands and algorithmic trading essentials, including argument types, NumPy basics for trading, and using pandas to import and scrape data.

Learn how to create and assign python variables, execute commands, store and retrieve values, and understand case sensitivity, including multiple assignments with x and y.

Learn how Python represents data types—integers, floats, strings, and booleans—with examples like 1145, -502, and 3.0, and check a variable’s type using the type command.

Learn how to type strings in python with single or double quotes, verify with the type function, print outputs, and concatenate strings with variables, including casting between strings and numbers.

Learn to group data by using pandas and compute statistics such as mean, median, standard deviation, and describe by groups like position or team, using an NBA dataset.

Explore time series visualization with Plotly for financial analysis, plotting interactive charts in three lines of code, and build TradingView–style graphs for algorithmic trading.

Explore Plotly’s data visualization tools in Python, building charts from scatter, plot, and histogram to ohlc and candle charts for market data; learn efficient, client-facing plotting techniques.

Explore scatterplot graphs to visualize time-series market data with Plotly, building interactive charts using x and y axes, markers, and sliders, illustrated by Apple stock data.

Explore candlestick charts with open, high, low, and close data in Plotly, and learn to update layouts, add annotations, and shapes to highlight periods like bull markets.

Master plotting open-high-low-close charts with a date axis, showing open, high, low, and close prices, and combine with a scatter line for mixed diagrams in Plotly.

Learn to access alternative data from AAD using an API key and pull historical market data. Retrieve stock prices, exchange data, sentiment analysis, and central bank data for diverse markets.

Explore the basics of the financial market in this generative AI course, including BlackRock and the SEC, and trading styles like string and pyramidal trading, taught by Semi.

Discover the basics of trading by exploring market players, trading strategies, and financial products; build your own diversified and competitive portfolio through fundamental and technical analysis, and algorithmic optimization.

Explore how stock, bond, currency, commodity, and cryptocurrency markets offer diversified investment options and identify the right securities for a balanced portfolio.

Explore generative AI models and the transformer architecture, learn their limitations and solutions, and practice with OpenAI and Gemini APIs, including sentiment analysis and market trend use cases.

Learn the essentials of neural language models, including the transformer architecture with self-attention and encoder-decoder structure, and how pre-training, fine-tuning, prompting, and retrieval augmented generation address limitations in finance.

Explore transformer fundamentals, including tokenization, embedding, and positional encoding, and learn how self-attention, multi-head attention, and the encoder-decoder architecture handle long sequences in natural language processing.

Learn how large language models pre-train on vast unstructured data from the internet and code, then fine-tune for domain tasks. Discover instruction using prompt engineering and retrieval augmented generation.

Fine-tune large language models to boost performance in specific domains and tasks using domain data or proprietary data, with full or parameter-efficient methods like LoRA.

Explore prompt engineering to guide llms with prompts, using zero-shot and few-shot learning with one, two, or n-shot examples, enabling precise sentiment classification and instruction-based outputs.

Explore retrieval augmented generation, combining retrieval and a generator to supply up-to-date, domain-specific or task-specific knowledge to LLMs, via embeddings, vector store, and cosine similarity, reducing hallucinations.

Explore open source and closed source LLMs, including llama 2, mistral, falcon, palm, gpt-3.5 turbo, and gpt-4, and examine parameters, tokens, and context windows.