The amount of data you need depends both on the complexity of your problem and on the complexity of your chosen algorithm.

The amount of data required for machine learning depends on many factors, such as:

• The complexity of the problem, nominally the unknown underlying function that best relates your input variables to the output variable.
• The complexity of the learning algorithm, nominally the algorithm used to inductively learn the unknown underlying mapping function from specific examples.

Generally, it is common knowledge that too little training data results in a poor approximation. An over-constrained model will underfit the small training dataset, whereas an under-constrained model, in turn, will likely overfit the training data, both resulting in poor performance. Too little test data will result in an optimistic and high variance estimation of model performance.

Impact of Dataset Size on Deep Learning Model Skill And Performance Estimates https://machinelearningmastery.com/impact-of-dataset-size-on-deep-learning-model-skill-and-performance-estimates/

How Much Training Data is Required for Machine Learning?

https://machinelearningmastery.com/much-training-data-required-machine-learning/

The answer to the question is not straight-forward since many factors including size of features, model parameter and experimental setup all play a role. Normally, there is a preference to have more training samples to realize a rich representation of the class of interest in the classification problem.

https://stats.stackexchange.com/questions/164500/is-this-training-dataset-enough-for-training-and-testing-classification-model?noredirect=1&lq=1

https://stats.stackexchange.com/questions/51490/how-large-a-training-set-is-needed?noredirect=1&lq=1

Good question. You can work on 100 records as minimum number of records but when you use data mining techniques you should work on bigger number to get better results, more data are better absolutely and this is why we use data mining algorithms. In general, more data mean more experience which will help to increase the classification accuracy and lower errors.

there is no boundary in this case it is good if you re working with big data like 10,000 thousand to 1 million which is ensure your accurate result .big data give you good accurate result so work with big data as many as data possible.

For understanding its concepts, you start from small & basic example, via few points via hand calculations or via MATLAB. Number of points is not a restriction.

Your questions aim directly to the core of all AI & ML methods. The answer to the least amount of data necessary to carry out reliable research has no unique answer.

It is worth to mention hypothesis creation and testing during each ML-based research, see details in the paper.

It depends on both: data, and the applied ML method. When you are collecting data to some known problem, you already have in your hands the answer about the least amount of data to get reliable results.

When you are carrying out new research on a new topic, it is advisable to test several ML methods and compare their results.

When you have a sufficient amount of data, it is advisable to test the dependence of your result robustness on the number of data!

Very similar testing was carried on in ML part of the paper on the prediction of heart arrhythmias from ECG recordings. (When interested, check my profile for the paper. It contains a description of the whole research cycle.)

Hi Motunrayo, as my experience the more amount of dataset you have, the more accurately it will predict. But recently I trained a decision tree and random forest classifier on my dataset which consists of 117 data samples only and it worked well. I'll say above 100 data samples will work well. All the best.

Hope this table can help. It really depends on your problem setting such as the feature size and underlying relationship.