Machine learning is transforming how we handle data, but the complexity of implementing algorithms can be daunting. Luckily, Easy Machine Learning simplifies this process with an innovative, dataflow-based system. This blog will guide you step-by-step on how to get started with Easy Machine Learning and troubleshoot common issues along the way.

What is Easy Machine Learning?

Easy Machine Learning offers a streamlined approach to applying machine learning algorithms, especially across distributed platforms like Hadoop and Spark. The system uses a directed acyclic graph (DAG) to illustrate the flow and processing of tasks. This involves multiple nodes where operations are represented, making it easier to visualize and manage workflows.

Getting Started: Setup Instructions

To set up Easy Machine Learning on your local machine, follow these steps:

• Clone the project repository.
• Prepare the necessary environments and development utilities.
• Refer to the instructions in Quick-start.md to create our system on your computer.

Using the Easy Machine Learning Studio

Once you’ve set up Easy ML, log in to the studio at localhost using the official account credentials:

• Username: bdaict@hotmail.com
• Password: bdaict

For a better experience, it is recommended to use the Chrome browser.

Creating a Machine Learning Task

After logging in, users can create a machine learning task (a dataflow DAG) using algorithms and datasets from the left panel:

• Select algorithms and datasets from the **Program** and **Data** panels.
• Clone existing tasks from the **Job** panel if needed.
• Configure task information and parameter values on the right panel.
• Task nodes can represent either a standalone program or a distributed one.

Submit your task by clicking the **submit** button. The status of each node will be represented with different colors, indicating progress.

Viewing Outputs and Logs

After completing a task, right-click on the **green output port** of any finished node to preview the output data. You can also check the execution logs:

• Right-click the node and select **Show STDOUT** or **Show STDERR** to view standard output or error logs respectively.

Modifying and Resubmitting Tasks

Even after finishing tasks, you can still modify and resubmit them. The system intelligently schedules only the affected nodes, reusing output from unaffected nodes to save time and resources.

Uploading Algorithms and Datasets

You have the option to upload your own algorithms and datasets to create custom tasks. Here’s how:

• Click on the **upload program** button and specify the algorithm package details.
• Make sure to define the command line pattern string accurately for inputs, outputs, and parameters.
• Similarly, you can use the **upload data** button for datasets.

How to Experience Our System

You can explore Easy Machine Learning through our online service:

• Outside ICT: http://159.226.40.104:18080/dev
• Inside ICT: http://10.60.0.50:18080/dev

For assistance or feedback while using the system, feel free to contact us at bdaict@hotmail.com.

Troubleshooting Common Issues

If you encounter difficulties while using Easy Machine Learning, consider the following troubleshooting tips:

• Ensure that your environment is correctly set up as per the instructions.
• Check the compatibility of the browser if you’re facing issues with the user interface—Chrome is recommended.
• Verify the accuracy of command line patterns while uploading algorithms. Use the helper tool provided in the panel to aid you.

For more insights, updates, or to collaborate on AI development projects, stay connected with fxis.ai.

At fxis.ai, we believe that such advancements are crucial for the future of AI, as they enable more comprehensive and effective solutions. Our team is continually exploring new methodologies to push the envelope in artificial intelligence, ensuring that our clients benefit from the latest technological innovations.