Using Neural Networks to Classify People

Question 1

In scientific literature, a neural network is demarcated as ‘a tremendously matching dispersed processor composed of guileless processing units that has an ordinary tendency for storing first-hand information and making it accessible for use. It looks a lot like the brain in two respects: (a) interneuron linking strengths (referred to as synaptic weights), which are used to store the attained information, (b) information is attained by the network from its environment through a learning process’ (Piasecki and Wojcicka-Wojtowicz 2017, 135).

Question 2

We can teach a neural network to be familiar with a particular pattern we are attracted to through either supervised learning or unsupervised learning or reinforcement learning (Shiffman, 2012). Through supervised learning, the teacher has to be smarter than the neural network. For instance, if we need to teach the network facial recognition, the network is shown different faces (but the teacher has to know the names of the faces). The network then makes its guesses while the person teaching it offers the correct answers (Shiffman, 2012). By comparing the correct answers with its answers, the network is then able to make adjustments. Afterwards, the network is able to recognize the specific pattern we are interested in.

On the other hand, the neural network can be taught through unsupervised learning. This teaching occurs when the person teaching the network does not know the answers to the data set (Shiffman, 2012). For instance, if you are interested in a specific hidden pattern, you can teach the network by hiding things until it finds what you are interested in. Conversely, through reinforcement learning, the neural network is taught through observation. The neural network makes a judgement with an outcome (turn left or right) and observes its environment (ouch or yum). In case the observation is negative, the network adjusts its weights so that it makes a different decision the next time (Shiffman, 2012).

Question 3

Three personal characteristics of a child (demographic) that can influence in the future adult aberrant behavior include (a) manipulative behaviors (b) low self-esteem, and (c) high dependency. Existing studies have found that child personal characteristics have lasting effects that form adult wellbeing. Child personal characteristics initiate and maintain certain life paths. However, adult welfare is the outcome of a multifaceted web of emotional, natural, and social influences re-counting over the life course, which makes it thought-provoking to trace the underlying paths linking child personal characteristics to adult outcomes (Hampson, 2008)

Question 4

I would recommend simulation, which is the use of analytical models for representing real processes. In simulation, one can apply tree of decision to track the above mentioned characteristics. This is because this software shows the several outcomes from a series of decisions. Through this tool, one can visualize all possible outcomes and decision points chronologically. For instance, with this tool, it can be easy to tract the characteristics of a child (the root), while growing (the branches) and when the child becomes an adult (the end of the decision tree). This will help understand and know if the child’s characteristics have any influence on the future adult. Software for building tree of decision include Decision Aide, Expert Choice, and Decision Pad (Serifi et al., 2007). Besides, a decision support system (DSS) assist people make more effective decisions by answering multifaceted questions. 

Get a brand-new guide on essay writing

Everything you need to get A+ this semester!

Get it

By continuing, you agree to the Terms And conditions and Privacy policy

The download will start within seconds. Good luck!

Download