OER·harvester

← Back to the library
Zenodo PDF resource

Artificial intelligence (AI) with It's Applications

The book "Artificial Intelligence (AI) with It's Applications" provides a comprehensive insight into the field of AI, exploring its fundamental principles, modern applications, and future potential. It serves as a valuable resource for students, researchers, and professionals looking to understand AI’s role in shaping industries and everyday life. The book begins with an introduction to Artificial Intelligence , cov…

Licence
OPEN CC-BY-4.0
Authors
Dr. Dipikaben Umakant Thakar, Mrs. PL. Natchiammai, Dr. R. J. Kavitha…
Published
2025-03-18 · Zenodo
Language
eng
Length
66700 words
Type
narrative text
Open ↗ Download Open original ↗

Use the melted DataFrame and specify x and hue

sb.countplot(x='value', hue='Class/ASD', data=df_melted)

plt.xticks(rotation=60)

plt.tight_layout()

plt.show()

Output

From the above plots we can draw the following observations,

Age_desc is the same for all the data points.

This used_app_before feature doesn’t seem to be useful or it will be the source of data leakage.

Here it seems like the chances of a male having autism is higher than a female but that is not true because we don’t have an equal number of examples of males and females.

plt.figure(figsize=(15,5))

sb.countplot(data=df, x='contry_of_res', hue='Class/ASD')

plt.xticks(rotation=90)

plt.show()

Output

In some places approximately 50% of the data available for that country have autism and in some places, this is quite low. This implies that the geography of a person also gives an idea of having autism.

plt.subplots(figsize=(15,5))

for i, col in enumerate(floats):

plt.subplot(1,2,i+1)

sb.distplot(df[col])

plt.tight_layout()

plt.show()

Output

Both of the continuous data are skewed left one is positive and the right one is negatively skewed.

plt.subplots(figsize=(15,5))

for i, col in enumerate(floats):

plt.subplot(1,2,i+1)

sb.boxplot(df[col])

plt.tight_layout()

plt.show()

Output

Ah! some outliers in the result column. Let’s remove that as it doesn’t seem like too much loss of information.

df = df[df['result']>-5]

df.shape

Output

(798, 22)

So, here we lost only two data points.

Feature Engineering

Feature Engineering helps to derive some valuable features from the existing ones. These extra features sometimes help in increasing the performance of the model significantly and certainly help to gain deeper insights into the data.