19.07.2017 - DAY 7 - NEURAL NETWORK

It is a good time to improve  skills. The road started aiming to construct building energy consumption program. It continues with deeper energy knowledge (themes, units, workflows, etc.) and mathematical equations such as neural networks and optimization algorithms. Day by day many professions goes through coding base artificial algoritms. Architecture discipline can not escape this general tendance. Directly it effects the discipline because of the collaborative workflow. As an personal idea, It was great decision to join MSTAS.2017. It was effective to learn computational discipline basics and understand my position as an architect because  my bachelor education constructed on design base education system because of the conventional architectural education background. This is not wrong but it needs to enlarge theoritical background and jump thhrough several new disicplines. I was complicated about learning coding and advanced energy learning, still sometimes think in a same way but MSTAS.17 shows that people waiting for innovation and risky attempts to move forward, to see new outcomes. You can see that my daily working preferences.

Before constructing ANN as program, I am planning to work with some grasshopper plugins that you can experiment ANN and optimization algortihms these are 'dodo' and 'crow'. These programs inventors also architects and they build this systems at master level it is good motivation for me to believe my aims. There is another possibility for ANN and energy, that is working with Matlab. Still it is decision right now but it is good to write at somewhere. For deep learning of Energy consumption as you can see I started to read Energyplus documents. EnergyPlus encourage architects and engineers to benefit from EnergyPlus database as I mentioned yesterday. For today that's all. Stay cool.

Links :


Popular posts from this blog

An Algorithm for Efficient Urban Building Energy Modeling and Simulation

07.09.2017 - How to create or modify EnergyPlus Weather Data (EPW)