Search This Blog

Simplicity is the ultimate sophistication.” — Leonardo da Vinci
Contact me: sreramk360@gmail.com

Friday, 23 March 2018

Forward designing vs backward designing

Forward designing vs backward designing


A task that involves intense planning and meticulous observation at each stage, can always be approached in two ways. You could either work backwards from your desired result and construct the solution, or you could work forwards by breaking down the problem and intuitively determining what requirements you will have to meet, and creating them before reaching the end result. I have been solving problems for as long as I remember. Over the past few years, I haven't just been solving problems that are atomic in nature (that cannot be broken down into multiple smaller problems), I have also been solving problems that can be broken down into multiple parts, where each of part must be solved separately. In all these years, there were two things I have been doing wrong. As a result, the amount of work and time investment I put in was not (to the degree I was expecting) aligned with the results I gained from them.

In forwarding designing, the designer starts with the resources and information he already has based on quick intuitive assessments. The designer forms a road-map that leads to the solution and breaks up the problem into multiple smaller problems and starts solving them. Usually, a lot of time goes into the design phase. In this development model, the entire production system or the corporate or individual building the system will not have any working model until all the problems are solved and assembled together. This entire process could consume a lot of time.

In backward designing, the designer thoroughly analyzes the end-solution systematically designs a working prototype as fast as possible. Using this prototype, the designer would visualize the final resultant working of the system in a very short period of time. But this system will not be treated as the final result, and it will be subject to further refinement or on many occasions, a total reconstruction will be done. Reconstructing the same product after having gone through the development process during the prototype phase would give the developing team an intuitive understanding of what could go wrong at each stage. This intuitive understanding will help designers optimize their production process.

The forward designing process involves constant reconstruction and allows ambiguous structuring of the development process. This gives way to innovation and lets designers explore new possibilities. The necessity to explore and innovate will be forced as no working product will be available until the entire development process completes. But the result produced by the backward designing process will be more likely to be aligned with the actual requirements and to meet the assigned time limit. Forward designing process sacrifice predictability and structuring for a creative outcome. But the backward designing process is the exact opposite.

Saturday, 3 June 2017

Nature of logical statements and postulates.



Nature of logical statements and postulates.

Postulates are usually propositions that do not have a definite proof. They are usually statements concluded from an experiment. A Logical statement is defined as an assertion obtained by the application of rules that are defined to be correct. In the context of this article, term ‘rule’ is used interchangeably in place of the term ‘logical-statement’ and the term rule-set is always used interchangeably with the phrase a set of logical statements’. And by definition, a logical statement is a statement derived from fixed principles that are known to be correct. Contrary to the popular belief that every statement can be proved, in reality, not all statements can be proven. This is because, in the end, every rule is constructed by fixed rules and to prove the ‘fixed rules’ correctness, we will need more ‘fixed rules’.

Thursday, 15 December 2016

Neural Networks

Neural Network Library (work in progress) 

               -By K Sreram  

I have recently written a simple neural network algorithm and I have posted it on github. It uses feed-forward and back propagation algorithms. It is still far from a complete neural network tool. The link to the source: https://github.com/Aplokodika/MachineLearning/tree/master/src  


Featured post

Why complicating matters is not good

“ Simplicity is the ultimate sophistication.” — Leonardo da Vinci Why is complicating things wrong ? - K Sr...