MACHINE

LEARNING &
PREDICTIVE
ANALYTICS
Prese nta t i o n o n Ho usi ng Da ta
 OUR TEAM
• ABDUL AZHAR (N20231043)
• AVADHI BADJATYA (N20231319)
• GAURAV AGRAWAL (N20231061)
• ISHAN SHRIVASTAVA (N20231096)
• SHRUTI SINHA (N20231018)
• SAILESH KUMAR PATRA (N20231320)
Theoretical Relationships:
• The total height of a stacked bar represents the total number of
houses within that particular average household size range.
• The sum of the heights of all segments within a bar equals the
total number of houses in that household size range.
• There is no inherent order between the household size ranges
on the X-axis.

01 INTERPRETATION
By analyzing the heights and segments of the stacked bars, we can understand how the median housing value distribution
varies across different household sizes. For instance, a bar with a prominent green segment might indicate that a
significant portion of houses suitable for that household size range fall within the $100,000 - $200,000 median housing
value range.
Theoretical Relationships:
• The total height of a stacked bar represents the total number of
blocks with a specific range of total households.
• The sum of the heights of all segments within a bar equals the
total number of blocks in that total household range category.
• There is no inherent order between the total household range
categories on the X-axis.

02 INTERPRETATION
Overall, this theoretical representation provides a framework to understand the information encoded within the stacked
bar chart. By analyzing the heights and segments of the stacked bars, we can gain insights into how the total number of
households in a block is distributed across different block categories. This information can be helpful for understanding
demographics and housing trends in block communities.
Additional Insights:
• You can compare the distribution of blocks across different average
household count ranges. For instance, the chart might reveal that
blocks with a smaller average number of households (0-500) are
less frequent compared to blocks with a moderate household range
(e.g., 500-1500).
• This information can be valuable for urban planners, sociologists,
or real estate agents who study demographics and housing trends in
block communities.

Distribution of Blocks by Average Number of Households pen_spark

03 INTERPRETATION
· Distribution of Blocks by Average Household Count: This chart allows you to visualize how the average number of
households in a block is distributed across different block categories.
· Example: Let's look at the bar for the average household range of 1000-1500. The green segment might be the
tallest, indicating that a higher number of blocks fall within this range compared to blocks with a lower or higher average
household count.
Additional Insights:
• X-Axis (Horizontal): This axis represents the ocean proximity of the housing
units, likely categorized as (text labels on the X-axis):
• oceanfront
• Near Ocean
• Near Bay
• Island
• Inland
• Y-Axis (Vertical): This axis represents the number of housing units (unit
specified in the chart, e.g., thousands). The scale is likely shown on the left
side of the Y-axis.

Stacked Bar Chart

04 INTERPRETATION
• Distribution of Housing Units by Vacancy and Ocean Proximity: This chart allows you to visualize how vacancy
status (occupied or vacant) is distributed across different categories of ocean proximity for housing units.
• Example: Let's look at the bar for "Near Ocean" proximity. The blue segment (occupied) might be taller than the
green segment (vacant), indicating that a higher number of housing units near the ocean are occupied compared to
vacant ones.
05 • There is a positive correlation between median house value
and median income. This means that houses in
neighborhoods with higher median incomes tend to be more
expensive.
• There is a positive correlation between population and
households. This means that areas with more people tend to
have more households.
• There is a negative correlation between median house value
and housing median age. This means that older houses tend
to be less expensive.

• The value in the row labeled "total_bedrooms" and the column labeled "avgBedrooms" is 0.74. This means that there is a strong
positive correlation between the total number of bedrooms in a house and the average number of bedrooms in a house. This is
expected, since a house with a higher total number of bedrooms is also likely to have a higher average number of bedrooms per
room.
06 The green line represents the number of rooms in a block, and
the blue line represents the number of rooms in a block
altogether.
• The x-axis represents the total number of rooms in a block.
• The y-axis represents the number of houses.
The graph suggests that there are more houses with a lower
number of rooms in a block. As the number of rooms increases,
the number of houses decreases.

This suggests a right-skewed distribution, where most of the data points fall on the left side of the graph.
The green line (average) is likely above the blue line (median) throughout the graph, which could indicate that the average number of
rooms in a block is higher than the median number of rooms. This can happen in a right-skewed distribution.
07 The x-axis represents the average number of rooms in a house,
ranging from 0 to 10 rooms. The y-axis represents the number
of houses.
Here’s a theoretical representation of the graph:
• The x-axis represents the average number of rooms in a
house. There are likely equal intervals between each
number on the x-axis.

• The y-axis represents the number of houses that have a corresponding average number of rooms. The tick marks on the y-axis
are not labeled but likely increase by constant intervals.
The graph suggests that the number of houses follows a bell-shaped curve distribution around a central value. This type of
distribution is symmetrical, where the left half of the graph mirrors the right half. The most common number of rooms likely falls
around the center of the x-axis, with the number of houses gradually decreasing as the number of rooms deviates from this central
value in either direction.
08 The y-axis likely represents the probability density of a house
having a specific number of rooms. Probability density is a
function that describes the probability of a random variable
falling within a certain range of values. In this case, the y-axis
value at a specific number of rooms on the x-axis represents the
probability density of a house having that exact number of
rooms. The total area under the curve would equal 1, which
represents the total

probability of a house having any number of rooms between 2,000 and 12,000 rooms (given the range of the x-axis).

• The graph suggests that the distribution of the number of rooms is skewed to the right. This means that most houses have a lower
number of rooms in a block. The probability density appears to be highest around 4,000 rooms and then decreases as the number
of rooms increases. There are also likely to be a few houses with a very high number of rooms.
• The green line represents the average number of rooms in a block. The location of the green line intersects the curve on the x-
axis at the average number of rooms.
 Housing Data - Neural Network Analysis

• The output of this code snippet is primarily the test Mean Squared Error (MSE), which indicates
how well the neural network model is performing on unseen data. The MSE represents the
average squared difference between the actual and predicted values. Lower MSE values indicate
better performance, as they suggest that the model's predictions are closer to the actual values.
• In the context of regression tasks like predicting median house values, the MSE provides insight
into the accuracy of the model's predictions. A low MSE indicates that the model can make
accurate predictions, which is crucial for tasks such as housing price prediction, where small errors
can have significant consequences.
• In summary, the output significance lies in the evaluation of the model's performance on unseen
data, indicating how well the model generalizes to new observations.