Thursday, July 13, 2017
Monday, July 10, 2017
I've got a series of results by a deep neural networks classifier after 25 hour run on a program this morning. This time, I had set an automatic mode for its running so that I enjoyed a good sleep while the laptop cared of the run process. ImageJ, that is a bio image converter, shows a good result of the classification. Anyway, I go on to analysis on the results.
Saturday, July 8, 2017
The backpropagatioin program still runs in the process of classification on my laptop since last night. It crushed in the mid night, and I had rebooted it on the same machine that has eight threads on an chip. What a time consuming program this is!
Friday, July 7, 2017
A classifier program still runs on my new laptop! How many hours does this deep learning classifier need to complete the task of training and classification? We have great challenges for improvement of this approach that is surely promising a solution. Anyhow I cannot join an evening mess at my institute, because I cannot move away from my lab.
A series of severe rain has washed Tokyo and cleaned up everything during these two days at night. I am running a deep belief net on a new laptop that I set up just now. Everything looks going well as written in papers by the U of Tronto Team. I'm thinking in which point my swarm dynamics algorithm works and betters off.
Sunday, June 18, 2017
The rainy season has begun at the beginning of this June, and no rainy days continue in cool air since then in Tokyo. This evening is the true beginning of rainy season. It is alike three years ago. Anyway, this year is very much easy to live.