Umhlahlandlela Wokuhlaziya Ukusinda NgePython: Ukusebenzisa Amamodeli Esikhathi Somcimbi Ukubikezela Isikhathi Sokuphila Kwekhasimende

# Time 
median_survival = kmf.median_survival_time_
print(f"Median Customer Lifetime: {median_survival} months")

Singakwazi futhi ukwenza okunye ukuhlaziya, njengokuqhathanisa phakathi kwamaqembu. Cabanga ukuthi le nkampani yakwaTelco ihlukanisa amakhasimende ayo ngamaqembu amabili:

1. Abasebenzisi abanzima: Frequency of Use > median
2. Abasebenzisi abathambile: Frequency of Use <= median

Singakwazi ukuqhathanisa kokubili imisebenzi yokusinda kulawa maqembu amabili.

# Column Groups
df['Heavy_User'] = np.where(df['Frequency of use'] > df['Frequency of use'].median(), 1, 0)
df.head()

plt.figure(figsize=(12, 5))
plt.title('Kaplan-Meier Survival Curve: Telco Customer Lifetime')
plt.xlabel('Time (months)')
plt.ylabel('Probability Churn')

# Fit the model for Soft users and plot
kmf.fit(df[df.Heavy_User == 0]['Subscription  Length'], df[df.Heavy_User == 0]['Churn'], label='Soft User')
ax = kmf.plot_survival_function()

# Fit the model for Heavy users and plot
kmf.fit(df[df.Heavy_User == 1]['Subscription  Length'], df[df.Heavy_User == 1]['Churn'], label='Heavy User')
ax = kmf.plot_survival_function(ax=ax)

plt.show()

Futhi kukhona. Nakuba abasebenzisi abakhulu behlala beqinile nenkampani kuso sonke isikhathi esibekelwe, abasebenzisi abathambile bazonyakaza ngokushesha ngemva kwenyanga engama-30. Isikhathi sabo esimaphakathi sokusinda yizinyanga ezingama-40.

Uma uqhathanisa amaqembu, kufanele uqiniseke ukuthi umehluko ubalulekile ngokwezibalo. Ngalokho, iphakheji lifelines kusetshenziswe ukuhlolwa kwezinga lelogi. Kuyisivivinyo se-hypothesis:

• Ho (null hypothesis): Amajika okusinda wabantu ababili awahlukani.
• Ha (alternative hypothesis): Amajika okusinda kwabantu ababili ahlukene.

from lifelines.statistics import logrank_test
# 3. Perform the Log-Rank Test
results = logrank_test(df[df.Heavy_User == 0]['Subscription  Length'],
                       df[df.Heavy_User == 1]['Subscription  Length'],
                       event_observed_A= df[df.Heavy_User == 0]['Churn'], 
                       event_observed_B= df[df.Heavy_User == 1]['Churn'])

# 4. Print Results
print(f"P-value: {results.p_value}")
print(f"Test Statistic: {results.test_statistic}")

if results.p_value < 0.05:
    print("Result: Statistically significant difference between groups.")
else:
    print("Result: No significant difference detected.")

P-value: 7.23487469906141e-103
Test Statistic: 463.7794219211866
Result: Statistically significant difference between groups.

Ukusebenzisa i-Cox Proportional Hazard

Into yokuqala epholile ongayenza nge-Cox Proportional Hazard [CPH] Imodeli ibheka ukuthi ezinye izinto eziguquguqukayo zingaba nomthelela kanjani ekusindeni komuntu wakho ombhekile.

Asihlephule.

1. Siqala ngokukhetha amanye ama-covariate
2. Sihlunga idathasethi
3. Faka imodeli
4. Faka imodeli

# 1. Prepare the data
# Selecting the time, the event, and our chosen covariates
cols_to_use = [
    'Subscription  Length', # Time 
    'Churn',                 # Event (E)
    'Charge  Amount',        # Covariate 1
    'Complains',             # Covariate 2
    'Frequency of use'       # Covariate 3
]

# Dropping any missing values for the model
df_model = df[cols_to_use].dropna()

# 2. Initialize and fit the Cox model
# Use the penalizer to stabilize the math if not converging.
cph = CoxPHFitter(penalizer=0.1)
cph.fit(df_model, 
        duration_col='Subscription  Length', 
        event_col='Churn')

# 3. Display the results
cph.print_summary()

# 4. Visualize the influence of covariates
cph.plot()

Lona umphumela wethu omuhle.

Singakuhumusha kanjani lokhu?

Umugqa oqondile onedeshi kokuthi 0.0 iphuzu elingathathi hlangothi.

• Uma iphuzu eliguquguqukayo lihlala 0akunawo umthelela ku-churn.
• Kwesokudla (> 0): Yandisa ingozi (yenza ukushuba kwenzeke ngokushesha).
• Kwesokunxele (< 0): Yehlisa ubungozi (yenza ikhasimende lihlale isikhathi eside).
• Etafuleni, ikholomu ebaluleke kakhulu yababambe iqhaza ebhizinisini i-Hazard Ration exp(coef). Lisitshela ukuthi isiphindaphinda umphumela engcupheni ye-churn.

[TABLE] Izikhalo (5.36): Ikhasimende elikhonondayo 5.36 izikhathi (noma 436%) amathuba amaningi ukuxova nganoma yisiphi isikhathi kunekhasimende elingakhonondi. Lona umphumela omkhulu.

[GRAPHIC] Uyakhononda (Ingozi Ephezulu): Lesi isibikezelo sethu esiqine kakhulu. Amakhasimende anezikhalazo cishe 5.4 izikhathi ezingaphezulu ukuqhuqha nganoma yisiphi isikhathi uma kuqhathaniswa nalabo abangakwenzi.

[TABLE] Imvamisa yokusetshenziswa (0.99): Nakuba i-p-value ithi lokhu kubalulekile ngokobuchwepheshe, i-HR ka-0.99 ngempumelelo ingu-1. Kusho ukuthi umthelela ku-churn awunaki (ushintsho olu-1% kuphela).

[GRAPHIC] Imvamisa Yokusebenzisa (Ukungathathi hlangothi): Isikwele sihlezi cishe ngqo kulayini ongu-0.0. Kule modeli ethile, ukuthi ikhasimende liyisebenzisa kaningi kangakanani isevisi akushintshi kakhulu nini bayaqhuma.

[TABLE] Inani Lenkokhiso (0.83): Kukho konke ukukhuphuka kweyunithi eyodwa, ingcuphe ye-churn yehle ngo-17% ($1 – 0.83 = 0.17$). Amakhasimende akhokha kakhulu azinzile.

[GRAPHIC] Inani Lenkokhelo (Isici Esivikelayo): Isikwele singakwesokunxele kumugqa onguziro. Amanani aphezulu ahlotshaniswa ne-a ngaphansi ingozi yokutheleleka.

Singaphinde sibheke kokubili imisebenzi ye-Survival kanye ne-Hazard yale modeli.

Ijika lifana nemodeli ye-KM. Ake siqhathanise amathuba okusinda enyangeni yama-34 efanayo.

# Extract the baseline survival probability at time 34
survival_at_34 = cph.baseline_survival_.loc[34]
print(f"Baseline Survival Probability at period 34: {survival_at_34.values[0]:.4f}")

Baseline Survival Probability at period 34: 0.9294

Cishe iphakeme ngo-3%, ngo-~93%

Futhi ukuvala lesi sihloko, ake sikhethe amakhasimende amabili ahlukene, elilodwa elingenazikhalo nelinye elinezikhalazo, futhi ake siqhathanise amathuba awo okusinda enyangeni yama-34.

# 1. Pick a customer (or predict for a new one)
individual = df_model.iloc[[110,111]]

# 2. Predict their full survival curve
pred_survival = cph.predict_survival_function(individual)

# 3. Get the value at time 34
prob110_at_34 = pred_survival.loc[34].values[0]
prob111_at_34 = pred_survival.loc[34].values[1]

print(f"Customer 110 (no complaints) Probability of 'Surviving' to period 34: {prob110_at_34:.2%}")
print(f"Customer 111 (yes compaints) Probability of 'Surviving' to period 34: {prob111_at_34:.2%}")

Customer 110 (no complaints) Probability of 'Surviving' to period 34: 93.94%
Customer 111 (yes compaints) Probability of 'Surviving' to period 34: 61.68%

Umehluko omkhulu, huh? Ngaphezu kwama-30%. Futhi ekugcineni singakwazi ukubala isikhathi ezinyangeni lapho ikhasimende ngalinye kulindeleke ukuthi lisebenze.

# Time Until Churn (Expected life) by customer
pred_churn = cph.predict_expectation(df_model.iloc[[110,111]])

# Get the values in months
prob110_churn = pred_churn.loc[110]
prob111_churn = pred_churn.loc[111]

print(f"Customer 110 (no complaints) expected churn at: {prob110_churn: .0f} months")
print(f"Customer 111 (yes compaints)  expected churn at: {prob111_churn:.0f} months")

Customer 110 (no complaints) expected churn at:  41 months
Customer 111 (yes compaints)  expected churn at: 31 months

Impela, izikhalo zenza umehluko kule nkampani yakwaTelco.

Ngaphambi kokuthi Uhambe

Nokho, ukuhlaziya ukusinda kungaphezu nje komsebenzi wezibalo. Izinkampani zingayisebenzisa ukuqonda ukuziphatha kwamakhasimende.

Amamodeli e-Kaplan-Meier kanye ne-Cox Proportional Hazard ahlinzeka ngemininingwane engenzeka empilweni ende yababhalisile. Sibonile ukuthi okuguquguqukayo okufana nenani lekhasimende kanye nezikhalazo zesevisi kuyithinta kanjani ngokuqondile i-churn, okuvumela abenzi bezinqumo ukuthi baphishekele amasu okugcina aqondiswe kakhulu.

Ochwepheshe bedatha abaqondayo lawa mamodeli bangakha ithuluzi elinamandla lezinkampani ukuthuthukisa ubudlelwano bazo nesisekelo sazo sabasebenzisi. Sebenzisa lawa mathuluzi ukuze uhlale ngaphambi kwejika. Ngokwezwi nezwi.

Uma ukuthandile lokhu okuqukethwe, ngithole kuwebhusayithi yami.

I-GitHub Repository

Izithenjwa

[1. Survival Analysis Definition] (

[2. The Complete Introduction to Survival Analysis in Python] (

[3. Introduction to Customer Survival Analysis: Understanding Customer Lifetimes] (

[4. Ultimate Guide to Survival Analysis] (

[5. What is the difference between Kaplan-Meier (KM) and Cox Proportional Hazards (CPH) ratio?] (

[6. Lifelines Documentation] (

[7. Survival Analysis in R For Beginners] (