I've come up with a new method of fitting that fits a function to a given date. I have described the mathematics of it and proved relevant theorems to show how it works. The only meaning of this method is its application for machine learning. When I go to mathematicians, they say the theorems are correct, but the mathematics of them is not unusual, but expected. They do not know much about machine learning and have no inclination to know. It is therefore difficult to impress a mathematical journal editor for publication acceptance. Mathematicians therefore recommend turning to machine learning experts. On the other hand, when I go to machine learning experts, they are reluctant to comment because they do not readily understand the relevant math (unless they need some time and refer to some books). Moreover, the concept is somewhat counterpointed to the latest beliefs in the machine learning world, where almost everyone believes that the ML problems need to be solved in very large dimensions, and most successful tools such as deep convolutional neural networks or the Traditional core methods or the graph-based methods are designed taking into account the high dimensions and the conviction that the ML problems can only be solved in very large dimensions. My methods require a solution as small as possible, which requires a traditional representation of domain knowledge-based features in combination with dimension reduction tools as a preprocessor to reduce the dimensions. So when I talk to ML people about my method, they might say that ML problems can best be solved in very large dimensions, so my method is impractical for very high dimensions.

I can apply and demonstrate my method to solve some ML records like IRIS. I can also illustrate the inner workings of my method by visualizing simulated datasets in 2 dimensions, for illustrative purposes only.

I need better workstations and some time and resources to apply for and solve harder ML problems. For that I need support and resources. This is only possible if someone buys my idea and my sponsor as a startup type. My strategy is to first publish this mathematical method of functional fitting in a mathematical journal in order to maintain some authenticity and to help me get serious attention from ML experts for providing laboratories / infrastructure or to attract venture capitalists to a startup ,

I am happy to suggest that my strategy is a good idea. If so, which mathematical journals can I focus on for this purpose? I do not expect to go to mathematicians and say that I did something incredible, but I just want to arouse enough interest to be published in a Descent Journal, so it's easy for me to grab the attention of the ML world to win.